Hydrogen production: Perspectives, separation with

Journal of Industrial and Engineering Chemistry 49 (2017) 1–25

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

Review

Hydrogen production: Perspectives, separation with special emphasis on kinetics of WGS reaction: A state-of-the-art review Samrand Saeidia,b , Farhad Fazlollahic , Sara Najarid, Davood Iranshahib , Jirí Jaromír Klemeše,* , Larry L. Baxterf a Department of Wood and Paper Science, Research Institute of Forests and Rangelands, Agricultural Research, Education and Extension Organization (AREEO), Tehran, Iran b School of Chemical Engineering, Amirkabir University of Technology (Tehran Polytechnic), No. 424, Hafez Avenue, 15914 Tehran, Iran c School of Chemical Engineering, Purdue University, 480 Stadium Mall Dr., West Lafayette, IN 47906, USA d Department of Chemical Engineering, Tarbiat Modares University, Tehran 14115-114, Iran e Pázmány Péter Catholic University, 1088 Budapest, Szentkirályi utca 28, Hungary f Chemical Engineering Department, Brigham Young University, Provo, UT 84602, USA

A R T I C L E I N F O

Article history: Received 3 July 2016 Received in revised form 1 December 2016 Accepted 1 December 2016 Available online 15 December 2016 Keywords: Water–gas shift reaction Pd-alloy membrane Reaction kinetic models Membrane reactor Hydrogen production

A B S T R A C T

Fossil fuel depletion, global warming, energy security, and climate change spur interest in commercial and environmentally friendly alternative fuels. Palladium-based catalytic membrane technology currently produces ultrapure hydrogen from fossil fuels. Palladium exhibits high permeability, selectivity for hydrogen, and good surface properties. Properties of some palladium alloys enable the industrial production of hydrogen. Reaction rates and conversion depend on several parameters. This document reviews kinetic expressions for the water–gas shift (WGS) reactions and WGS combined with Fischer– Tropsch synthesis (WGS–FTS) reactions at high and low temperatures along with the details of the catalysts and operating conditions. The discussion includes mathematical reactor modeling. © 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evolution of WGS reaction and the WGS–FTS kinetic rates . . . . . . . . . . . . . . High-temperature WGS reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Low temperature for WGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . WGS kinetic models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Micro kinetic models for WGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macro-kinetic models for WGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fischer–Tropsch synthesis kinetic models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Micro mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . High-temperature WGS in FTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Low-temperature WGS in FTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macro kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . High-temperature WGS in FTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Low-temperature WGS in FTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H2 perm-selective membrane reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Palladium-based membrane reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Membrane reactor types used for hydrogenation and dehydrogenation reactions Packed bed membrane reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fixed-bed membrane reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .. ..

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

* Corresponding author. http://dx.doi.org/10.1016/j.jiec.2016.12.003 1226-086X/© 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

2 4 4 5 5 6 7 11 11 12 12 14 14 15 15 15 16 16 16

2

S. Saeidi et al. / Journal of Industrial and Engineering Chemistry 49 (2017) 1–25

Fluidized bed hydrogen perm-selective membrane reactors Alternatives for hydrogen production reactions . . . . . . . . . . . . . Water–gas shift process (WGS) . . . . . . . . . . . . . . . . . . . . . . Steam reforming process . . . . . . . . . . . . . . . . . . . . . . . . . . . Dehydrogenation process . . . . . . . . . . . . . . . . . . . . . . . . . . . Pd-membrane development . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tubular reactor models for WGS in FTS . . . . . . . . . . . . . . . . . . . Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . . .

Introduction The majority of the world’s energy demand is supplied by fossil fuels. The CO2 produced from these fuels contribute to climate change [1,2], which is the primary reason for shifting focus to clean and renewable energies [3]. This goal requires a revolution in energy production, conversion, storage, and distribution technologies [4]. Hydrogen plays an important role in some of the alternative fuel proposals and is sometimes proposed as an alternative fuel, but it is more accurately characterized as an energy carrier since it is not a primary energy source [5]. Similar to electricity, hydrogen is not a usable natural source of energy but must be generated from other sources. Hydrogen could play a significant role in the future power generation and transportation energy sectors [6]. Hydrogen as the lightest element with 7% the density of air and is an odorless, tasteless and colorless gas [7]. The energy yield of hydrogen oxidation is about 122 kJ/g, which is 2.75 times greater than hydrocarbon fuels on a mass, but not on a volume, basis [8]. Application of hydrogen in transportation systems, whether as a combustion engine fuel or in a fuel cell, receives some favorable attention as an energy policy issue [9]. Hydrogen utilization generates few pollutants and little CO2 directly [10], though hydrogen production can be quite CO2 and pollutant intensive. Vehicles fueled by hydrogen might decrease dependence on fossil fuel and mitigate tailpipe emissions [11], depending on how hydrogen is produced. The efficiency of hydrogen fuel cell (HFC) vehicles can be up to three times higher than gasoline engines [12]. However, hydrogen presents greater storage and transportation barriers than exist for the liquid fuels. The hydrogen production process poses economic and environmental challenges [13,14]. Compared to methane (CH4) and propane (C3H8) energy densities of 32.6 and 86.7 MJ/m3 (at 15 C and 1 atm), respectively, the energy density of hydrogen is 10 MJ/m3, which means that hydrogen in vehicles requires pressurized or other compacted storage and usually a large fuel tank [15]. Moreover, hydrogen is prone to leak. Free radicals from form from hydrogen reaction with ultraviolet radiation, leading to ozone depletion [16]. Conventional approaches for H2 production sometimes involve syngas generation from hydrocarbons produced from fossil-fuel resources, commonly natural gas, followed by the WGS (water–gas shift) reaction to convert CO and water to H2 and CO2 [17]. H2 and CO form aliphatic hydrocarbons through Fischer–Tropsch or similar processes or sometimes alcohols or ethers through other processes [18]. Alternatively, electrolysis can produce hydrogen from water. Based upon previous works from the literature [19– 23], the advantages and disadvantages of different hydrogen production techniques are outlined in Table 1. Currently, 48% of the total demand for hydrogen is produced from steam reforming of natural gas, about 30% from heavy oils and naphtha reforming, 18% from coal gasification, 3.9% from water electrolysis and 0.1% from biomass and other resources [24]. Consequently, steam reforming

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

16 17 17 17 18 19 20 22 23 23 23

deserves special attention with emphasis on the kinetics of WGS reaction (Fig. 1). The WGS reactor was first discussed in 1888 [25], and it became well-known with the Haber ammonia synthesis process and the development of catalyst in 1912 [26]. The catalyst comprised of iron and chromium reduced CO content to about 2% at 4000– 5000 C. The WGS reaction is an exothermic, reversible reaction which can be expressed as Reaction Eq. (1). COðgÞ þ H2 OðgÞ $CO2ðgÞ þ H2ðgÞ

DH 0298 ¼ 41:09 kJ=mol

ð1Þ

Increasing temperature shifts the equilibrium constant of this reaction toward reactants (H2O and CO) and increases the reaction rate. Since the total gas volume does not change, pressure effects are mainly second-order non-idealities. Metals and metal oxides can catalyze the WGS reaction. Iron oxide– chromium oxide was used as a catalyst in ammonia plants in an adiabatic single reactor and produced CO at 2–4% while the temperature increased along the length of the reactor. This is near the equilibrium composition at these conditions [26]. The iron oxide/chromium oxide catalysts works best at high temperatures, and this process is known as high-temperature, shift catalysts (HTSC). More recently, copper-based catalysts operate at temperatures as low as 473 K and achieve CO concentrations of 0.1–0.3% in these reactors. The temperature limitation primarily occurs because of the gas mixture due point at pressure. These copper-based formulations are low-temperature shift catalysts (LTSC). The kinetics of the water–gas shift reaction as a side reaction in FTS is also important. FT synthesis proceeds through a complex reaction mechanism involving many species. WGS kinetic mechanisms describe carbon monoxide depletion rates and carbon dioxide formation. FTS converts synthesis gas in a product spectrum consisting of a complex, multi-component mixture of linear and branched hydrocarbons and oxygenated products. The main products are linear paraffin’s and a-olefins. Water is a primary product of the FT reaction, and CO2 can only be produced by the water–gas shift (WGS) reaction (RWGS = RCO2 ). The WGS reaction is a reversible, parallel–consecutive reaction with respect to CO [27]. WGS reactions occur simultaneously with hydrocarbon production in some FTS reactions. The conventional reactor for FTS is a tubular reactor. WGS reactions occur in two adiabatic stages including high- and lowtemperature shifts having inter cooling to maintain the inlet temperatures [26]. Pasel et al. have proposed employing isothermal reactors based on their experiments [28]. Since the copperbased catalyst is prone to poisoning by sulfur compounds originating from coal or hydrocarbon sources, this configuration is needed; whilst the iron-based catalyst is more sulfur tolerant. A guard bed eliminates the sulfur compounds for protecting the catalyst if the feed contains sulfur. This review analyzes the high- and low-temperature reaction rates and mechanisms in the form of kinetic expressions related to

S. Saeidi et al. / Journal of Industrial and Engineering Chemistry 49 (2017) 1–25 Table 1 Technical pros and cons of potential methods of H2 generation [5]. H2 generation techniques Conventional thermochemical technologies

Gasification

Pyrolysis

Reforming

Approaches Coal gasification

Industrialized technology, relatively inexpensive feedstock, almost energy intensive, less tar and char production than that in pyrolysis, higher efficiency for steam gasification than air/oxygen gasification Coal pyrolysis Industrialized technology, almost high gas efficiency, relatively inexpensive feedstock, ability for tar recovery, less energy intensive and less expensive than coal gasification, the most inexpensive conventional H2 production process Industrialized technology, the highest efficiency among Steam methane the other conventional methods, almost energy intensive reforming due to high temperature requirement (SMR) Partial Industrial technology, almost energy intensive due to high oxidation temperature requirement Auto-thermal reforming

Conventional electrochemical technologies

Bio-hydrogen thermochemical technologies

Thermochemical cycle

Water sulphur– iodine cycle

Electrolysis

Water electrolysis

Cons Not environment friendly due to CO2 by-product emission, low-quality hydrogen

Not clean and environment friendly, production of a considerable amount of char, reactor clogging

Not environment friendly due to CO2 by-product emission, dependence on fossil fuel (methane)

Not environment friendly due to CO2 by-product emission, dependence on fossil fuel (methane), high oxygen supply, less efficient than SMR Industrial technology, almost energy intensive due to high Not environment friendly due to CO2 by-product temperature requirement, less capital costs emission, dependence on fossil fuel (methane), less efficient than SMR Emission free due to being a completely closed cycle, Very high temperature requirement, highly corrosive alternate source, suitable for continuous operation, due to using reactants such as iodine and sulphuric acid, suitable for using with nuclear, solar and hybrid heat high capital costs due to need for advanced materials for sources process construction

Industrialized technology, emission free

Low efficiency, expensive due to having a high capital cost, usable in areas with inexpensive electricity cost

Bio-catalysed electrolysis Cold/hot Plasma Plasma reforming/ gasification PhotoPhotoelectrochemical catalytic water splitting

Clean and renewable, alternate source

Low efficiency, expensive due to having high capital cost

No harmful emissions, industrialized (in the case of plasma gasification) technology, high energy density, good process control Clean and renewable, alternate source, applicable using either artificial or natural light

High electrical power consumption, high operating costs, necessitating occasional maintenance

Pyrolysis

Fast pyrolysis

Low efficiency (Gas product = 10-25 mol%), formation of tar and char

Gasification

Steam gasification

H2 recovery from gaseous and liquid products, operating temperature (T = 500 C) lower than both combustion and gasification, ability to take place with the absolute absence of oxygen High-purity and high-quality H2 (Gas product = 85%), operating temperature lower than partial oxidation (T = 800 C) Reasonable hydrogen yield, 30-50% higher fuel yield per unit feed than conventional biomass gasification, operating at lower temperatures with higher efficiencies compared to conventional gasification, easy ash and tar removal, low CO2 emission, continuous process H2 generation from wet biomass such as sludge, slurry, municipal solid waste, rich in H2

Solar gasification

Supercritical water gasification Biological/ Biochemical technologies

Pros

Photosynthetic

Direct biophotolysis

Indirect biophotolysis

Fermentative

Photofermentation

Dark fermentation

Operating at ambient temperatures and pressures, less energy-intensive, environment friendly, H2 generation from water using solar, 10-folds increase in solar energy conversion than trees and crops Operating at ambient temperatures and pressures, less energy-intensive, H2 generation from water, environment friendly, N2 fixation from atmosphere, circumventing the inhibition of H2 evolving by O2 due to separating O2 and H2, ability to combine microalgal H2 generation with wastewater treatment Operating at ambient temperatures and pressures, less energy-intensive, environment friendly, lack of O2 evolving activity, ability to use wide spectrum of light, flexibility to use various organic wastes as substrates, less inexpensive energy generation than photolysis processes Operating at ambient temperatures and pressures, less energy-intensive, very clean and environment friendly, anaerobic phenomena, H2 generation without using light energy, flexibility to use various organic wastes as substrates, production of valuable by-products such as butyric acid and lactic acid, less inexpensive energy generation than photolysis processes

Low efficiency, kinetically very slow reaction, expensive due to having high capital cost, suffering from catalyst decay and recombination

Energy intensive, tar formation, lower gasification efficiency at higher moisture content Requiring efficient and expensive solar collector plates, weather dependant, costly

In the early stages of technical development, costly

Costly raw material, low hydrogen yield and rate of hydrogen production, strong inhibition effect of generated oxygen on hydrogenase enzyme, no waste employment Costly raw material, lower hydrogen yield and rate of hydrogen production, economically high-risk long-term process

Costly raw material, lower hydrogen yield and rate of hydrogen production, using nitrogenase enzyme with high energy demand, low solar energy conversion efficiency, needing large areas for the anaerobic photo bioreactors, low light conversion efficiency Costly raw material, lower hydrogen yield and rate of hydrogen production, needing CO2 separation

3

4


Table 1 (Continued) H2 generation techniques

Bio-hydrogen Electrical discharge technologies

Hot plasma

Approaches

Pros

Cons

Sequential dark and photofermentation

Rapid H2 production, simple operation due to operating at Costly raw material, low hydrogen yield and rate of ambient temperatures and pressures, less energyhydrogen production, difficult process in large scale due intensive, environment friendly, higher H2 yield than to requiring continuous light source single dark or photo-fermentations, production of Metabolites with ability to H2 conversion

Plasma gasification/ pyrolysis

Low harmful emissions, industrialized (in the case of High electrical power consumption, high operating plasma gasification) technology, high energy density, good costs, necessitating occasional maintenance process control, flexibility to mix different types of biomass such as municipal solid waste, tires and hazardous waste, almost no production of leachable bottom fly ash/ash

Export Export steam steam

Heat recovery

ZnS

NG

Process Process steam steam

Pre-reforming (Optional)

Desulphurization

Steam reforming

Fuel Fuel gas gas Demineralized Demineralized water water

WGS

CO2

CO2 capture

Hydrogen purification (PSA)

Pure Pure H H22

H H22 Slipstream Slipstream

Tail-gas Tail-gas

Tail-gas Tail-gas Fig. 1. Steam reforming of natural gas.

WGS and WGS–FTS reactions. The discussion summarizes the evolution of independent WGS and WGS–FTS kinetics in terms of catalyst composition, catalyst size [29], operating conditions, catalyst life time, feed ratio and kinetic model type. The second section discusses hydrogen perm-selective membrane reactors with particular focus on palladium-based membranes. Finally, the paper discusses some innovations to improve the performance of perm-selective membrane reactors. All of this information is summarized in several tables. The discussion focuses on mathematical modeling of tubular FTS reactors. The effect of the kinetic rate on reactor configuration is highlighted. Evolution of WGS reaction and the WGS–FTS kinetic rates Figs. 2 and 3 illustrate the number of high- and lowtemperature and macro- and micro-kinetic publications from

1977–2015. Finally, Fig. 4 shows the trend the recent publications related to these reactions for hydrogen and fuel production. High-temperature WGS reactions The high-temperature ferrochrome catalysts operate in the range of 580–725 K [30]. The WGS reaction frequently operates adiabatically at industrial scale where the temperature increases along the reactor length. A modest inlet temperature of 623 K lowers the bed temperature sufficiently to prevent catalyst damage. This inlet temperature maintains a maximum bed temperature of approximately 823 K at the exit. Newsome reports a typical high-temperature catalyst composition of 74.2% Fe2O3, 10.0% Cr2O3, 0.2% MgO, with the balance of volatile materials, where chromium oxide with an optimal content of 14% acts as a stabilizer which prevents the iron oxide from sintering [31]. The 8%

120 Journal Publicaons

100 High-Temperature

80

Low -Temperature

60 40 20 2014

2012

2010

2008

2006

2004

2002

2000

1998

1996

1994

1992

1990

1988

1986

0

Year Fig. 2. Distribution of published papers for low and high-temperature from year 1977 to 2015.

Fig. 3. Distribution of published papers for Micro and Macro mechanism model from 1977 to 2015.

S. Saeidi et al. / Journal of Industrial and Engineering Chemistry 49 (2017) 1–25 Table 2 Typical catalyst compositions for low-temperature shift.

140 120

Number of Publicaons

5

100 80

Components

Percent

Ref.

Components

percent

Ref.

ZnO CuO Cr2O3 Mn, Al and Mg oxides

68–73% 15–20% 9–14% 2–5%

[31]

CuO ZnO Al2O3

32–33% 34–53% 15–33%

[25,34]

60 40 20 0

Year Fig. 4. Number of WGS and WGS-FTS journal papers from 1977 to 2015.

Cr2O3 helps prevent sintering or other surface area degradation [32]. A high-temperature reactor can react to equilibrium concentration at 723 K (about 3% CO). Industrial-scale reactors operate from atmospheric pressure to 83 bar and with CO concentrations ranging from 3 to 80% [31]. The pretreatment of HTSC activates the catalyst [25] and typically partially reduces Fe2O3 (hematite) to Fe3O4 (magnetite) using process gas mixtures, as Reactions (2) and (3) show. This pretreatment also converts any hexavalent CrO3 to Cr2O3 [33]. 3Fe2 O3ðsÞ þ H2ðgÞ ! 2Fe3 O4ðsÞ þ H2 OðgÞ DH0298 ¼ 16:3 kJ=mol

ð2Þ

3Fe2 O3ðsÞ þ COðgÞ ! 2Fe3 O4ðsÞ þ CO2ðgÞ DH 0298 ¼ þ24:8 kJ=mol ð3Þ The gas feed mixture concentrations should fall in a range to prevent over reduction. A reduction factor proposed by Rhodes [32] controls the reduction process. The typical value of the reduction factor (R) should be less than 1.2 and values more than 1.6 represent over reduction [32,33]. ½CO þ ½H2 ð4Þ R¼ ½CO2 þ ½H2 O The ratio of steam to CO is an important parameter in hightemperature shift reactions. When reaction occurs at low ratios of steam to CO, metallic iron, methanation, carbon deposition and FT reactions occur [26]. Callaghan et al. suggest a contact time of approximately 3–9 s for the reaction [34]. The literature frequently suggests new catalyst compositions. Grenoble and Estadt [35] analyzed the catalytic effect of alumina, silica and carbon coated with groups VIIB, VIII and IB metals. They reported that the metal or metal oxide should be acidic. The turnover number decreases in the order of Cu, Re, Co, Ru, Ni, Pt, Os, Au, Fe, Pd, Rh and Ir supported on alumina. In addition, commercial cobalt–molybdenum oxides catalyze the reaction above 623 K and are sulfur tolerant [36]. Higher conversions were reported for the high-temperature catalyst using different promoters [37–39]. Rhodes and Hutchings [40] involved the addition of 2 wt.% of B, Pb, Cu, Ba, Ag and Hg metals to the high-temperature catalyst and revealing the promoter effect on the conversion. The activity of the promoters is reported as Hg > Ag, Ba > Cu > Pb > un-promoted > B at the temperatures between 623 to 713 K. The successfully promotion of the iron-chromia catalyst by Cu, Au and Ru has can also promote chromia catalysts been reported [40]. Low temperature for WGS The low-temperature shift reaction occurs at 473–523 K on a mixture of ZnO, CuO and Cr2O3/Al2O3 as a catalyst with varying composition [41] (Table 2).

Recent catalysts can also be operated at medium temperatures of around 573 K. The active species in the catalyst is the copper metal crystallites, which is more susceptible to thermal sintering and hence should not be operated at higher temperature [26]. ZnO and Cr2O3 provide the structural support for the catalyst and Al2O3 is largely inactive and helps in the dispersion and minimizes pellet shrinkage [33]. The lower operating temperature depends on the dew point of the mixture at industrial conditions. Rase proposes a catalyst from a mixture of sulfur and halogen. It is intolerant to unsaturated hydrocarbon and he recommends that it be protected from these compounds [42]. Twigg and Spencer add ZnO to sulfur catalysts because ZnO reduces sulfur poisoning of copper effectively [43]. Therefore, a guard bed of ZnO is used to prevent sulfur poisoning for the low-temperature shift reactor. The advantages of the low-temperature catalyst are its selectivity and increased hydrogen production with fewer side reactions occurring at the higher operating pressures. The normal lifetime of the low-temperature catalyst is 2–3 years [42]. The exit concentration of CO is 0.1% in the low-temperature reactor, which is desirable on commercial catalysts. Similar to the high-temperature catalyst, the low-temperature catalyst needs to be activated through exposure to the process stream with dilute H2 [25]. The CuO forms copper through the following reaction as the catalyst activates. 0

CuOðsÞ þ H2ðgÞ ! CuðsÞ þ H2 OðgÞ DH 298 ¼ 80:8 kJ=mol

ð5Þ

Since the CuO reduction reaction is exothermic, the catalyst should be exposed to the process stream at temperatures of 500 K or less as higher temperatures will lead to sintering [34]. Moreover, the process stream should contain hydrogen and steam vapor because liquid water could negatively affect the catalyst. Tanaka et al. indicate a mixture of CuMn2O4, CuAl2O4 and metal oxides has greater activity than the commercial catalyst [44]. Henrik Kusar find the copper ceria catalyst is non-pyrophoric and stable [45]. Dinesh studied Mn promoted Cu/Al2O3 catalyst and indicates that CO conversion up to 90% on catalysts with 8.55 wt.% Mn at 513 K and in 5.33 h space time [46]. Many more studies on various modifications in the catalysts are being proposed routinely in the literature. However, the mainstay of all the low-temperature catalysts still is copper. WGS kinetic models Newsome applied the Temkin model to low-temperature range shift reactions while Herwijnen and De Jong listed other kinetic models for the low-temperature water–gas shift Herwijnen and De Jong [47]. The models for WGS reaction at low temperature include the power-law model, the redox mechanism, and the model by Campbell and Moe based on industrial data. In addition, Ovesen et al. include a pressure correction in the power-law expression, while Rase extrapolates the experimental data that is normally conducted at the atmospheric pressure with the inclusion of a pressure correction formula and compares it with industrial data. Amadeo and Laborde analyze two redox models and three Langmuir–Hinshelwood (LH) models [48]. They found that the LH model was most accurate because it includes the adsorption of the four components as well as the surface reactions. They also

6


CO2 HO CO

OH

O C

CO2

H

H

O C O H

OH CO

H

CO2

OH

H 2O Fig. 5. Water–gas shift mechanism via formate intermediate.

modified the LH model based on the experimental result for the low-temperature shift reaction. Meanwhile, Sun found the LH model as the best predictive kinetic model for WGS reactions over precious metals [57]. Micro kinetic models for WGS Generally, the WGS reaction mechanisms utilizing metal oxide catalyst can be divided into two categories: Regenerative mechanisms Associative mechanisms The regenerative or redox mechanism postulates that the reaction occurs as a result of an oxidation reduction cycle on the catalyst surface. Eqs. (6) and (7) summarize the mechanism [49] H2O(g) + red ! H2(g) + ox

(6)

CO(g) + ox ! CO2(g) + red

(7)

First, H2O reduction on the catalyst surface produces H2. CO then oxidizes to CO2. This mechanisms explains the hightemperature water–gas shift reaction. Several authors explain the low-temperature shift reaction in terms of both the regenerative and associative mechanisms [50]. Ovesen et al. used a single copper crystal to investigate the surface redox mechanism and proposed an eight-step model. They found that there was more reaction on the reaction sensitive structure of Cu (11 0) than Cu (111) [51]. Wang investigated the surface redox mechanism at low temperatures using both the forward and reverse kinetics. The dissociation of water appeared to be the rate controlling step [52]. Rhodes and coworkers argued there is the possibility of the reduction reaction happening but they are doubtful about the possibility of the oxidative step [25]. The associative mechanism presents an alternative to the redox reaction. The adsorption–desorption model postulates that CO2 and H2 result from the decomposition of an intermediate that comes from the interaction of an adsorbed species. The aforementioned reaction is shown in [53] Eq. (8). CO(g) + H2(g) ! (intermediate) ! CO2(g) + H2(g)

(8)

Many authors cite this type of WGS reaction mechanism. However, there exists no definitive evidence about the nature of the intermediate involved in the associative mechanism. Formate and carboxyl intermediates represent the leading candidates. A Langmuir–Hinshelwood process produces Cu-Chromite, as mentioned by Rhodes. The reaction occurs either in the gas phase

or on the surface, where the reaction between either a hydroxyl species or water and carbon monoxide produces a formate species. The decomposition of water leads to the formation of the hydroxy intermediate, whereas the reduction of either adsorbed or gaseous carbon dioxide produces the formate intermediate (see Fig. 5). Rhodes [25] discovered that the intermediate was from the formate species. The involvement of both reactants and products of the WGS reaction in the principal decomposition pathway for formic acid led Grenoble and Estadt to suggest that formic acid was the intermediate for the process [35]. Ovesen et al. extended the redox mechanism based on the eight-step microkinetic model by the addition of 3 extra steps to account for formate, where they indicated that at atmospheric conditions, the formate coverage was unconsidered, but it was significant at higher pressures. Carbon monoxide oxidation and water dissociation represent the rate-limiting steps in this analysis [54]. Gideon Botes used SASOL data (Corporation of Oil and Gas) to evaluate three different mechanisms and determined that the best fit of the data involved the associative formate mechanism [55]. However, Waugh, supported the redox mechanism and questioned the formate mechanism after the author applied the micro kinetic approach [56]. Salmi et al. investigated the transient response of the WGS reaction and found that H2 was slowly liberated, which is inconsistent with the associative mechanism. They proposed a mechanism that involves both adsorptive and regenerative reaction steps [58]. The transient response investigations showed different responses when ferrochrome catalyst and Co–Mo catalysts were used [59]. Keiski investigated the high-temperature water–gas shift reaction using steady-state and transient techniques and determined that the rate determining steps in the transient investigations were CO adsorption, CO2 desorption and H2 formation [60]. Fishtik and Dutta determined that at higher temperatures, the reaction was governed by the redox mechanism while the associative and formate mechanisms were significant in the low-temperature region [61]. Cu (111) was used in the WGS reaction where Gokhale proposed the mechanism of carboxyl as the intermediate route for the reaction. In these interpretations, formate is non-reactive, and they observed a loss in activity with the increase in CO2 concentration due to the active sites that have been blocked at high pressure [62]. Mao concluded that the formate mechanisms is improbable based on density functional theory analyses and observed that copper catalyst increased the probability for the carboxyl and the redox mechanisms [63]. The carboxyl mechanism

S. Saeidi et al. / Journal of Industrial and Engineering Chemistry 49 (2017) 1–25 Table 3 The model of micro-kinetic for WGS reaction on Cu(111) [34,61].

Table 5 WGSR equilibrium constants [26].

Reaction number

Elementary reaction step

Ea (kJ/mol)

K0 desorption/ Adsorption (1/atms) Surface reaction(1/s)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

CO + S $ COS H2O + S $ H2OS H2OS + S $ OHS + HS COS + OS $ CO2S+ S COS + OHS $ HCOOS + S OHS + S $ OS + HS COS + OHS $ CO2S + HS HCOOS + S $ CO2S + HS HCOOS + OS $ CO2S + OHS H2OS + OS $ 2OHS H2OS + H.S $ OHS + H2S OHS + HS $ OS + H2S HCOOS + OHS $ CO2S + H2OS HCOOS + HS $ CO2S + H2S CO2S $ CO2 + S HS + HS $ H2S + S H2S $ H2 + S HS + HS $ H2 + 2S

0 0 106.27 44.769 0 64.852 0 5.858 17 121 110.04 5.439 3.766 61.086 22.18 64.015 23.01 64.015

1.5 106 1.0 106 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 1.0 1013 4.0 1012 1.0 1013 6.0 1012 6.0 1012

z¼

Macro-kinetic evaluations of the water–gas-shit reaction data frequently dome from its use in the ammonia process [67]. Table 4 lists the commonly used models. The Eley–Rideal and Langmuir– Hinshelwood (LH) type models generally represent the associative mechanism. Several approaches calculate the equilibrium constant for the WGS reaction at different temperatures. The equilibrium constant values in temperatures of interest appear in Table 5. Twigg listed

Table 4 WGS reaction kinetic expressions. High-temperature WGS [26]

Kodama et al. Hulburt–Vasan Bohlboro et al.

ð1þK CO ½COþK H2 O ½H2 OþK CO2 ½CO2 þK H2 ½H2 Þ2

½CO ½H k ½CO½H2 O 2 2 r ¼ ð1þK CO ½COþK H O ½H2 OþK COK ½CO2 þK H 2

2

2

½H2 Þ

k½H2 O r ¼ 1þK½H 2 O=½H2

r ¼ kPaCO PbH O PcCO 2

Oxidation reduction

½CO2 ½H2 K

2

PdH2

k1 k2 f½CO½H2 O½CO2 ½H2 =Kg r ¼ k1 ½COþk 2 ½H2 Oþk1 ½CO2 þk2 ½H2

Low-Temperature WGS [175] Kulkova and Temkin Campbell et al. Shchibrya et al. Goodgidge and Quazi

r ¼ kPCO r¼k r¼

P

H2 O

ð1 bÞ PCO PH2 O ð1bÞ

ð1þK CO PCO þK H2 O PH2 O þK CO2 PCO2 þK H2 PH2 Þ2

kPH2 O PCO ð1bÞ APH2 O þPCO2

r ¼ kPaCO PbH O PcCO PdH ð1 bÞ r ¼ kPCO PH O ð1 bÞ 2

Moe

0:5

PH2

2

2

2

523.15 83.956

573.15 38.833

673.15 20.303

673.15 11.723

723.15 7.3369

773.15 4.9035

1000 1 T

T ¼ Temperature ðKÞ

5693:5 þ 1:077lnðT Þ þ 5:44 104 T 1:125 ln K eq ¼ T 49170 13:148 107 T 2 T2

Macro-kinetic models for WGS

kK CO K H2 O ½CO½H2 O

473.15 210.82

823.15 3.4586

ð10Þ

The equilibrium constant can also be derived from thermodynamics [68] and is given in Eq. (11).

performed better than the formate mechanism when Grabow et al. used Pt (111) catalyst in the low-temperature water–gas shift reaction [64]. Tang investigated the WGS reaction on Cu (111) using the density functional method and a slab model and concluded that the carboxyl and redox mechanisms were more feasible than the associative and formate mechanisms [65]. There appears to be no consensus on a simple WGS mechanism. Hence, elementary step mechanisms may be necessary to describe the micro-kinetics of the WGS reaction [66]. Callaghan et al. [34] proposed such a mechanism on Cu (111) using 18 steps, as listed in Table 3.

r¼

T(K) Keq

the constants of equilibrium for the WGS reaction in the temperature range from 200 to 1199 C [26]. Twigg derived the following Eqs. (9) and (10) for the equilibrium constants ð9Þ K eq ¼ exp zðzð0:63508 0:29353z þ 4:1778Þ þ 0:31688

(X.S is the adsorbed species and S is the vacant site).

Langmuir–Hinshelwood

7

ð11Þ

Moehas derived a simple equation to employ in an empirical model which is given by Eq. (12) 4577:8 4:33 ð12Þ K eq ¼ exp T Tables 5 and 6 list the detailed kinetic expressions for both highand low-temperature catalysts, and Table 5 lists the kinetic parameters corresponding to the power-law kinetic expression. These parameters are grouped based on similar catalysts. The Arrhenius parameters obtained by Rhodes come from a series of experiments using the high-temperature catalysts as summarized in Table 6 [40]. Table 6 also contains the power-law reaction rates for low-temperature catalysts at different conditions. A comprehensive analysis of the previously published parameters was provided by Choi, Stenger and Koryabkin. The developed complex kinetic expressions and also the orders of the components from the power-law kinetic expressions describe the WGS reaction. Podolski and Kim that both the LH and power-law models reproduced their experimental data [69]. LH model also predicts the reaction rate in the presence of H2S. Rase [42] provides a kinetic expression for both the high- and lowtemperature shift reactions in his book which has been employed by Elnashaie and Elshishini [70] in their models. Singh and Saraf [71] extend the kinetics of laboratory low- and high-temperature catalysts to industrial scale using factors for diffusion limitation, catalyst lifetime, pressure correction and the effect of exposure to H2S. San et al. produce two rate expressions after carrying out the high-temperature reaction using different catalyst compositions. Keiski et al. [60] provide an expression for the high-temperature shift reaction. Recently, Thomas and Barton [72] use both the model of San et al. and that of Keiski et al. and also introduce the correction factor for porosity to model a heterogeneous reactor. Zhao et al. have investigated the effect of pressure by implementing the reduced rate method, and the LH model was proven to be better than the other models [73]. Choi and Stenger used Cu/ZnO/Al2O3 catalyst to determine the validity of different models (adsorptive, regenerative and empirical), and all the models gave good results with the best fit provided by the single step regenerative model. Table 7 lists the empirical equation that has been developed. In addition, the steam to CO ratio in the kinetic expression has been included by Wei and his coworkers. Nobel metal systems have not reached a mature commercial stage, with different combinations of newer catalysts regularly

8


Table 6 Power-law parameters for various WGS reaction catalysts and operating conditions. Catalyst properties

High-temperature CuO2CeO8O2-y(Cuceria) CuO1CeO9O2-y(Cuceria) Fe3O4/Cr2O3, 8 wt.% Cr2O3 Fe3O4/Cr2O3,180–250 mm Fe3O4/Cr2O3, 180–250 mm Fe3O4/Cr2O3, 180–250 mm Fe3O4/Cr2O3 Ru Ru/ceria Ni Ni/Ceria Rh Rh/Ceria 89% Fe2O3, 9% Cr2O3 80–90% Fe2O3, 8–13% Cr2O3, 1–2% CuO 80–95% Fe2O3, 5–10% Cr2O3, 1–5% CuO CuO/Fe3O4/Cr2O3, 180–250 mm CuO/Fe3O4/Cr2O3, 180–250 mm CuO/Fe3O4/Cr2O3, 180–250 mm Pd

Pd/Ceria Pt Pt/Ceria Power gas—pilotplantdata Girdler—pilotplant data ICIpilotplant data 0.4%Pt/Al2O3 Ce(La)Ox Low-temperature 5 at.% Cu–Ce (10% La) Ox 5 at.% Ni–Ce (10% La) Ox ICI – CuO–ZnO–Al2O3 ICI52-1(Copper basedcatalyst) density = 5.83 g/cm3

Operating conditions

Reaction order

Ref.

Ko

Ea (kJ/mol)

l

m

n

q

1.8 103(s1) 4.0 103 (s1) lnKo = 11.5 lnKo = 10.1 02 lnKo = 12.0 02 lnKo = 7.4 0.1 (lnKo = 26.1) 1.6 107(s1) 5.0 107(s1) 8.0 107(s1) 1.7 108(s1) 3.0 109(s1) 1.5 1010(s1) lnk0 = 14.78 102.845(mol/s gcat) 100.659(mol/s gcat) lnKo = 2.0 0.1 lnKo = 5.5 0.1 lnKo = 4.0 0.1 4.0 106(s1)

61 78 112 118 1 124 1 111 1 95 80 80 85 85 130 130 E/R = 9598 111 88 75 1 85 2 85 1 100

– – – – – – 1.1 – – – – – – 0.74 1 0.9 – – – –

– – – – – – 0.53 – – – – – – 0.47 0 0.31 – – – –

– – – – – – – – – – – – – 0.18 0.36 0.156 – – – –

– – – – – – – – – – – – – 0 0.09 0.05 – – – –

[45] [45] [32] [32] [32] [32] [60] [176] [176] [176] [176] [176] [176] [177] [178] [178] [40] [40] [40] [176]

4.0 107(s1) 1.0 106(s1) 2.5 10 7(s1) 9.4 107 (1/s) 1.47 108 (1/s) 4.5 109 (1/s) – –

100 80 80 89.5376 99.5792 112.968 39 58.5

– – – – – – 0.45 –

– – – – – – 0.37 –

– – – – – – 0 –

– – – – – – 0.73 –

[176] [176] [176] [42] [42] [42] [179] [180]

30.4 38.2 52.8 0.45

0 0 1 0.07

1 1 1 –

– – – –

– – – –

[180] [180] [181] [181]

1.07

0.55

–

–

– 200–250 mm, 393.15–523.15 K, 1:2CO/H2O 1 atm, 498.15–558.15 K 1 atm, 558.15–618.15 K 1 atm, 558.15 K 1 atm, 573.15 K 1 atm, 543.15 K 1 atm, 373.15 K 1 atm, 533.15 K 1 atm, 483.15–533.15 K 543.15 K 1 atm, 473.15 K 1 atm, 513.15 K 396–448 K, CO/H2O = 1/3

– – 3.99 106 (1/s) k = 5.37 107 (mol/m2s)/atml+m k = 4.40 105(mol/m2s)/atml + m – 2.96 105(1/s) (lnKo = 12.6) – – – – – – – – 1.9 106(molecules/s/site) – – 4.9 106(1/s)

41 47.4

0 1

1 1

– –

– –

[181] [181]

68 84 81 81 82 – 86 71 82.006 5.439 75 46 71

0.1 0.06 0.11 0.1 0.21 0.02 0.13 0.05 0.21 0.03 0 –

1.1 1 0.82 0.77 0.75 0.55 0.49 0.32 0.75 0.44 1 –

0.07 0.09 0.06 0.08 – – 0.12 0.85 – 0.09 – –

0.44 0.44 0.49 0.46 – 0.22 0.45 0.05 – 0.38 – –

[179] [179] [179] [179] [179] [179] [179] [179] [35] [179] [179] [45]

1 atm, 473.15 K 1 atm, 463.15 K 1 atm, 453.15–473.15 K – 1 atm, 403.15 K 1 atm, 473.15 K – 1 atm, 513.15 K 1 atm, 473.15 K 603.15 K 623.15 K 613.15 K 613.15 K 1 atm, 613.15 K 1 atm, 613.15 K

– – – – – – – – – 5.10 106(molecules/s/site) 3.23 105(molecules/s/ site) 1.18 105(molecules/s/site) 3.84 106(molecules/s/site) – –

67 79 86 69.3 55 62 55 56 32 96.232 5.439 95.395 10.46 79.914 3.347 106.69 5.858 71 42

0.2 0.8 1 1 0.3 0.9 1 0.9 0.7 0.1 0.24 0.08 0.13 0 0

0.6 0.8 1.4 1.9 0.38 0.8 1 0.4 0.6 0.44 0.53 0.69 0.35 0.5–1 1

0 0.9 0.7 – – 0.7 – 0.6 0.6 – – – – – –

0 0.9 0.9 – – 0.8 – 0.6 0.6 – – – – – –

[182] [182] [182] [181] [182] [182] [181] [182] [182] [35] [35] [35] [35] [182] [182]

473–623 K, CO/H2O = 1/3 573–623 K, CO/H2O = 1/3 1 atm, 623.15–713.15 K 1 bar, 653.15–723.15 K 6 bar, 653.15–723.15 K 27 bar, 653.15–723.15 K 3–5 bar, 846.15–906.15 K 573.15–1273.15 K 0.008–0.05 s contact time Coated on alumina support 5 wt.% loading

575–675 K 1 atm, 723.15 K, 6 mm 6 mm 1 atm, 723.15 K, 6 mm 6 mm 1 bar, 653.15–723.15 K 6 bar, 653.15–723.15 K 27 bar, 653.15–723.15 K 573.15–1273.15 K, 0.008–0.05 s contact time, Coated on alumina support 5 wt.% loading

1/4” 3/8”, 2.20 g/cm3 1/4” 1/4”, 1.25 g/cm3 11.3” 8.5 mm, 1.36 g/cm3 1 atm, 817.15 K 648.15–748.15 K

448.15–573.15 K, CO/H2O = 1.5 548.15–573.15 K, CO/H2O = 1.5 – 1 atm,473 K 1 atm,523.15 K

CuO–ZnO–Al2O3 Cu–ZnO Al2O3(EX-2248) SudChemie 1%Pt/Al2O3 1%Pt/Al2O3 1.66%Pt/Al2O3 1.66%Pt/Al2O3 2%Pt/Al2O3 0.9%Pt/Al2O3 1.4%Pt- 8.3% CeO2/Al2O3 2%Pt- 1%Re/CeO2ZrO2 Pt/Al2O3 1%Pt/CeO2 1%Pt/CeO2 42% CuO–ZnO–Al2O3, (G-66A), SudChemie Cu–ZnO–Al2O3 40% CuO–ZnOAl2O3 Cu–ZnO–Al2O3 Cu/Al2O3 10% Cu–Al2O3 8% CuO–Al2O3 CuO/MnO2 8% CuO–CeO2 8% CuO–15%CeO2–Al2O3 Rh/Al2O3 Rh/SiO2 Pt/SiO2 Pt/C Cu (111) Cu (11 0)

Arrhenius parameters

Table 7 WGS reaction macro kinetic rate expressions. High-temperature Catalyst properties

Operating conditions

72%Fe2O3–8%Cr2O3

Catalyst density = 4.561 g/cm3

Rate expression

Ref. 13

r ¼ Ef f 2:32 10 ðXCO

X CO Þexpð27760=RTÞ 3

K eq ¼ exp½ðð9998:22=TÞ 10:213 þ 2:7465 10

[183]

Ra Agf pf f s

T 0:453 106 T 2 0:201lnTÞ=R

LogAgf ¼ ð14:66 104 2 106 Þt

Girdler(G3-b)

Catalyst size = 0.62 cm equivalent diameter (1/4” 1/4”)

r ¼ kc

XC X D K ; k ¼ expð15:95 4900=TÞ 379rb

XA XB

c = 1.53 + 0.123 P for 11.8 < P 20 c = 0.816 + 0.184P for P 11.8 c = 4.0 for P > 20 SudChemie SHT-4

Catalyst density = , kg/m3

[185]

R = Universal gas constant (J/mol K) T = Temperature (K) d= Ratio of steam to CO

r ¼ k½CO0:9 ½H2 O0:25 ½CO2 0:6 ½H2 0 ð1 bÞ b ¼ ½CO2 ½H2 =K eq ½CO½H2 O, Ea = 114.6 kJ/mol r ¼ k½CO0:8 ½H2 O0:45 ½CO2 0:1 ½H2 0:1 ð1 bÞ

0.8–1.2 mm 603.15–773.15 K With H2S addition 653.15–773.15 K

For specific [H2S] l, m, n and q are power of CO,H2O,CO2 and H2 respectively ppmH2S 0 25 2000

8–13% Cr2O3 80–90 wt% Fe2O3,1–2% CuO 375–475 C

[70]

r = k(PCOPH2 O PCO2 PH2 /Keq) 2 k ¼ 1; 78 1022 1 þ 0:0097d 1:1364d T 8 exp 70=R T r = Reaction rate, Pi = Species partial pressure (kPa), K eq = Constant of equilibrium,

Ferrochrome catalysts Bohlboro Power Law Model

Ra = Relative activity, Agf = Aging factor [H2S] = H2S concentration t = Catalyst life time. Keq = Equilibrium Constant XCO = Mole fraction of CO in equilibrium condition fs = Rate reduction factor owing to concentration of H2S c = Factor of activity, r = Reaction rate (lbmol CO reacted/lb catalyst. hr), P = Pressure (atm), Xi = Concentration for component i (Ci/Cref), T = Temperature (K), K = Equilibrium constant, rb = Catalyst bulk density (lb/cu. ft),

For crushed Catalysts (Experimental)

r ¼ k½CO0:93 ½H2 O0:24 ½CO2 0:31 ½H2 0 ð1 bÞ Ea = 25.31071 kcal/mol

For coarse Catalysts (Commercial)

r ¼ k½CO0:87 ½H2 O0:26 ½CO2 0:18 ½H2 0 ð1 bÞ Ea = 59.8 kJ/mol

h

i 3 1 1 r ¼ ð3:2 1:3Þ 106 exp ð112 2Þ10 T 678 R

N 0.6 0.10 0.10

q 0 0.15 0.10

l 1 0.75–0.8 0.85

m [31] 0.25 0.50 0.40


pf ¼ Pð0:5P=250Þ f s ¼ 0:276logð½H2 S þ 2:78Þ þ 1:127 XCO ¼ XH2 X CO2 =X H2 O keq r = reaction rate (cm3/gcat h), XCO= CO mole fraction, R = gas constant (cal/gmol K), T = temperature (K), Eff = effectiveness factor, P = pressure (atm), Pf = pressure factor,

[186]

P1:17 0:12 P0:36 0:05 P0:09 0:05 P0:3 0:03 ð1 bÞ P0:84 0:04 CO H2 O CO2 H2 H2 S r = Reaction rate (mol/kg s), Pi = Species partial pressure (kPa), T = Temperature (K), b ¼ PCO2 PH2 =K eq PCO PH2 O

9

10

Table 7 (Continued) Low-temperature Model

Rate expression

Ref.

Model of Temkin

r¼ r = Reaction rate (1/s), k = Rate Constant (1/atm.s)

Model of Langmuir Hinshelwood

Girdler(G3-b)

O ½P CO P H

4577:8

[31]

1=atm:s Pi = Species partial pressure (Pa),

[187]

Cat 2 2 r60 , 3064 29364 12542 4577:8 6:74 18:45 , k ¼ exp 1:987T 1:987 þ 40:32 K eq ¼ exp T 4:33 , K CO ¼ exp 1:987T 1:987 K CO2 ¼ exp 1:987T 1:987 , 6216 K H2 O ¼ exp 1:987T 12:77 1:987 rcat = Catalyst density(gcat/cm3), r = Reaction rate(mol/cm3s), Pi = Species partial pressure (Pa), T = Temperature (K) r ¼ kPCO PH2 O ½1 ðPCO2 PH2 =K eq PCO PH2 O Þ,K eq ¼ exp 4577:8 4:33 k ¼ 1:85 105 exp 12:88 þ 1855:5 T T r = Reaction rate (mol/g min), Pi = Species partial pressure (bar) XC XD Catalyst particle size = 0.62 cm XA XB k ¼ expð12:88 2002:6=TÞ equivalent diameter (1/4” 1/4”) K r ¼ kc 379rb

r¼

ð1þK CO PCO þK H2 O PH2 O þK CO2 PCO2 Þ2

[188] [70]

c = 4.33 for P > 24.8 c = 0.86 + 0.14 P for P 24.8

Composition, of Catalyst CuO (32.7), ZnO (47) and Al2O3 (11)

1 atm, 453–503 K

P = Pressure (atm) r = Reaction rate (lbmol CO/lb cat. h), c = Factor of activity. Xi = Dimensionless concentration for component i (Ci/Cref), K = Equilibrium constant, T = Temperature (K), rb = Bulk density of catalyst (lb/cu. ft) 454:3 0:92 e PCO PH2 O ð1 bÞ T r¼

2 101:5 158:3 2737:9 1596:1 PCO þ 0:4e PH2 O þ 0:0047e PCO2 þ 0:05e PH2 1 þ 2:2e T T T T r = Reaction rate (mol/g min), T = Temperature (K),

Girdler/Sud Chemie (G66 b). CuO (32.2), ZnO(61.8) Fe2O3(1.6)

SudChemie (EX-2248) Cu/ZnO/Al2O3

PCO PH2 O 0.35–0.42 mm r ¼ 25:9 103 exp 16000 ð1þ127PCO PH2 O þ26PCO Þ RT particle size, Surface area22.3 (m2/g), pore diameter 12.5 nm r = Rate of reaction (mol/g s), Pi = Partial pressure (atm), 1 atm, 120 250 C, particle size of catalyst 200– 250 mm, H2: CO = 2

PCO PH PCO PH2 O K2eq 2 r ¼ 2:96 105 exp 47400 RT r = Reaction rate (mol/gcat h), Pi = Component partial pressure (atm),

ZnO (66%), CuO (33%)

5.09 g/cm3

[81]

Pi = Components partial pressure (Pa) b ¼ PCO2 PH2 =K eq PCO PH2 O [175]

R = Universal gas constant (J/mol K) T = Temperature (K) [181] T = Temperature (K), R = Universal gas constant (J/mol K)

r ¼ Ef f 2:955 1013 expð20960=RTÞ Agf Pt ðXCO X CO Þ LogAgf ¼ ð4:66 104 1:6 106 TÞt Pt ¼ Pð0:5P=250Þ XCO ¼ X H2 X CO2 =XH2 O keq Eff = Effectiveness factor Agf = Aging factor Pt = Effect of Pressure

[71]

X CO = CO mole fraction t = Age of catalyst in days, P = Pressure(atm)

r = Reaction rate (cm3/gcat h), XCO = Mole fraction of CO, T = Temperature (K)


Kinetic Model of Moe

kK CO K H

26800 ¼ exp T 4:33 ,k ¼ 6 1011 exp 1:987T 21500 9 A ¼ 2:5 10 exp 1:987T T = Temperature(K) Keq = Equilibrium Constant

kPH2 O PCO ð1bÞ APH2 O þPCO2 K eq

[189]

[57] q 0.9 0.7 0.9 m 1.4 1.5 1.9

2

2

Particle Size of catalyst = 3.3 mm Surface area = 80 m2/g Platinum

r¼

k PCO PH2 O

CO2 H2 K eq

DH2 = 11 kJ/mol, DH1 = 46 kJ/mol, A1 = 1 10 A = 9 107 mol/m3 atm2 s, Ea = 44 kJ/mol h i P P

ð1þK CO PCO þK H2 O PH2 O þK CO2 PCO2 þK H2 PH2 Þ K CO2 = 0.036 (1/bar), = 2.222(1/bar) K H2 = 2.197 105(1/bar), K H2 O = 2.006 105 (1/bar)

1 3 1

P H2 RT

1

PCO

1þA2 exp

RT

2

DH PCO PH 2 Oð1bÞ DH

1þA1 exp

– Precious metals

Catalyst Cu/ZnO/Al2O3 Cu/ZnO/Al2O3 Cu/Al2O3 a r ¼ Aexp E RT

g q n r ¼ Aexp ðEa=RT ÞPlCO Pm H2 O P CO2 P H2 P tot ð1 bÞ b ¼ PCO2 PH2 =K eq PCO PH2 O g = fudge factor correcting total pressure dependence

surface area = 10 m /g

2

For Large Catalysts 4.5 *4.5 mm, Bohlboro Cu/ZnO/Al2O3 40% Cu, 22%Zn, 5%Al

r ¼ k½PCO 0:8 ½PH2 O 0:5 ½PCO2 0:15 ½PH2 0 ð1 bÞ

2

2

Fischer–Tropsch synthesis kinetic models s , A2 = 43 s ,

Ea (kcal/mol) 20.674 18.69 14.173 P (bar) 5 20 20

l 1 1 1

d = Ratio of steam to CO Keq = Equilibrium constant, R = Universal gas constant (J/mol K) T = Temperature (K) r = Reaction rate, Pi = partial pressure (Pa),

. r ¼ kðPCO PH2 O Þ PCO2 PH2 K eq 2 k ¼ 1:74 1017 1 0:1540d þ 0:008d T 8:5 expð35=RTÞ SudChemie MDC 7 Cu-Zn based catalyst

11

reported in the literature. Platinum is one of the noble metal that is very promising and the kinetic expressions derived for platinum has been reported by Ding and Chan. The rate is exponential with frequency factor as follows: r ¼ Ea q n kP1CO P m H O P CO PH ð1 bÞ rate constant k ¼ k0 exp RT , b = the approach to equilibrium which given byb ¼ PCO2 PH2 =K eq PCO PH2 O and Ea = activation energy in terms of kJ/mol [74]. High- and low-temperature catalysts have similar kinetic expressions. Most experiments with different catalysts best fit the Langmuir–Hinshelwood and the power-law models. The recent investigations employ the San and Keiski kinetic expressions for high-temperature and Stenger and Choi for low-temperature WGS reaction. Mass transfer limitations, presence of impurities, experiments carried out at atmospheric pressure and the employment of an integral reactor for kinetic studies affect different opinions regarding the nature of the kinetics for the WGS reaction [75]. Rase corrected for pressure by taking into consideration of the diffusional effects of the catalysts, whereas Singh and Saraf determined the pressure correction factor.

n 0.7 0.7 1.4

[51]

[189]

[185]


FTS converts syngas (CO + H2) usually derived from natural gas, biomass, or coal to a complex, multi component mixture of linear and branched hydrocarbons and oxygenated products and is a well-established technology [76,77]. Several metals such as nickel, cobalt, ruthenium and iron have been shown to be active for this reaction. However, only iron and cobalt based catalysts appear to be economically feasible in an industrial scale. In the cobalt-based FTS, the oxygen from CO dissociation is almost completely discarded as water. Cobalt is not very active for the water–gas shift reaction: thus, in contrast to most iron-based Fischer–Tropsch catalysts, only a small fraction of the water produced is subsequently converted to carbon dioxide [78,79]. Iron-based catalysts provide both FTS and WGS reaction activities [80]. The FTS and WGS reactions can be shown as: COðgÞ þ ð1 þ n=2ÞH2 ðgÞ ! CHnðgÞ þ H2 OðgÞ

ð13Þ

COðgÞ þ H2 OðgÞ $CO2ðgÞ þ H2ðgÞ

ð14Þ

Where n is the average H/C ratio of the produced hydrocarbons. The WGS reaction is a reversible parallel-consecutive reaction with respect to CO and assumed that carbon dioxide is essentially formed by this reaction [76,77]. Micro mechanism Several mechanisms for the WGS reaction are proposed in the literature. Single studies of the WGS reaction over supported metals suggest the appearance of formate species. Formate intermediate mechanism is mentioned previously as shown in Fig. 5. Mechanistic structure of the water–gas shift reaction in the vicinity of chemical equilibrium. Furthermore, formate intermediate mechanism is the generally accepted mechanism of WGS reaction over iron catalysts in literature. All formate intermediate mechanisms are presented in Table 8. For derivation of the Langmuir–Hinshelwood–Hougen–Watson (LHHW) rate expressions, several assumptions were made as following: the WGS and FTS reactions proceed on different active sites, steady state for the adsorbed species, one rate-determining step (RDS) in the sequence of elementary WGS reactions [81], surface concentrations of intermediate species are negligible over the whole range of experimental condition and adsorption of

12


Table 8 Formate intermediate mechanisms for WGS of FTS. Model

Number

Elementary reaction

Ref.

WGS-I

1 2 3 4 5 6

CO þ s2 $COs2 CO2 þ s2 $CO2 s2 H2 O þ s2 $H2 Os2 H2 þ 2s2 $2Hs2 COs2 þ H2 Os2 $HCOOs2 þ Hs2 HCOOs2 þ s2 $Hs2 þ CO2 s2

[27]

WGS-II

1 2 3 4 5 6 7

CO þ s2 $COs2 CO2 þ s2 $CO2 s2 H2 O þ s2 $H2 Os2 H2 Os2 þ s2 $OHs2 þ Hs2 H2 þ 2s2 $2Hs2 COs2 þ OHs2 $HCOOs2 þ s2 HCOOs2 þ s2 $Hs2 þ CO2 s2

[27]

WGS-III

1 2 3 4 5 6

CO þ s$COs CO2 þ s$CO2 s H2 O þ s$H2 Os H2 þ 2s$2Hs COs þ H2 Os$HCOOs þ Hs HCOOs þ s$Hs þ CO2 s

[77,86,55]

WGS-IV

1 2 3 4 5 6 7

CO þ s$COs CO2 þ s$CO2 s H2 O þ s$H2 Os H2 Os þ s$OHs þ Hs H2 þ 2s$2Hs COs þ OHs$HCOOs þ s HCOOs þ s$Hs þ CO2 s

[77,86,55,85]

WGS-V

1 2 3 4 5 6 7 8

CO þ s$COs CO2 þ s$CO2 s H2 O þ s$H2 Os H2 Os þ s$OHs þ Hs H2 þ 2s$2Hs COs þ Hs$CHOs þ s CHOs þ HOs$HCOOs þ Hs HCOOs þ s$Hs þ CO2 s

[77,86,55]

WGS- VI

1 2 3 4 5 6

CO þ s$COs CO2 þ s$CO2 s H2 O þ 2s$OHs þ Hs OHs þ s$Os þ Hs COs þ Os$CO2 s þ s CO2 s þ s$CO2 þ s

[77,82]

WGS-VII

1 2 3 4

COs þ HOs$HCOOs HCOOs$CO2 þ Hs H2 þ 2s$2Hs H2 Os þ 2s$HOs þ Hs

[86,87]

1 2 3 4 5

CO þ s$COs H2 O þ s$H2 Os COs þ H2 Os$COOHs þ Hs COOHs$CO2 þ Hs H2 þ 2s$2Hs

[86,85]

WGS-VIII

reactants and desorption of products are at equilibrium. Also the surface concentration of the intermediates that take part in the rate-determining reaction is much higher than that of the other intermediates [77,82]. High-temperature WGS in FTS Nakhaei Pour et al. investigated Langmuir–Hinshelwood– Hougen–Watson (LHHW) kinetic models for the WGS reaction systemically based on detailed mechanism over an industrial Fe/ Cu/La/Si catalyst under FTS reaction conditions. They used four models to fit the experimental data over a wide range of reaction condition. WGS rate expressions based on the formate mechanism best described the WGS kinetic data (WGS III, WGS IV, WGSV and

WGS VI in Table 8). The WGS reaction rate was optimized with the kinetic expressions in [83] Table 9. In these rate equations, Pj is the partial pressure of species j in the effluent stream and K p is the equilibrium constant of the WGS reaction [83]. The corresponding model parameters appear in Table 10. In this table, the parameters of K1, K3, and K relate to adsorption coefficients of CO, H2O, and hydroxyl groups, respectively, for WGS III5, WGS IV6, WGS IV6, WGSV7 andWGSV7 expressions. They concluded that the basicity of the catalyst surface may not affect the WGS mechanism under FTS condition on the iron catalyst [77]. Chang et al. derived a series of rival models for WGS reaction using LHHW approach. They assumed that the rate of WGS reaction is controlled by desorption of CO2 via formate intermediate mechanism and studied comprehensive kinetics of slurry phase FTS on an industrial Fe.Cu.K.SiO2 catalyst using a stirred tank slurry reactor [84]. Their results are illustrated in Tables 8–10 [85]. Teng et al. studied the kinetics of water–gas shift (WGS) reaction over a Fe–Mn catalyst under Fischer–Tropsch synthesis (FTS) reaction conditions in a spinning basket reactor. The formate mechanism fits these experimental data better than the direct oxidation mechanism over the Fe–Mn catalyst under the FTS reaction conditions [82]. Yang et al. studied the detailed kinetics of the FTS over an industrial Fe–Mn catalyst in a continuous integral fixed-bed reactor. They derived reaction rate equations on the basis of the Langmuir–Hinshelwood–Hougen–Watson models for the Fischer– Tropsch reactions and the WGS reaction. They used formate species rate expressions due to a limited change of oxidation states of the iron cations [86]. Results of experiments are shown in Tables 9–11 [86]. Lox et al. investigated a detailed kinetic model using Hougen– Watson rate expressions for the Fischer–Tropsch reactions and the water–gas shift reaction on a precipitated promoted iron catalyst. The rate expression for the WGS employs elementary reactions involving a formate surface intermediate. The two-site reaction describing the formation of the formate intermediate is rate determining. Some sets of elementary reactions, rate expressions and kinetic parameters are shown in Tables 8–10 [87]. Low-temperature WGS in FTS Van der Laan et al. made several assumptions to derive the LHHW rate expressions for two cases, which are presented in Table 8. They developed two kinetic rate equations based on a formate mechanism. The expressions are given in Table 9. Table 10 gives the optimized values of the parameters in these two models: WGS-II6, WGS-I5. They obtained kinetics of the gas–solid FTS over a commercial Fe–Cu–K–SiO2 catalyst in a continuous spinning basket reactor [27,41]. Wang et al. achieved rate expression for a WGS reaction over an industrial Fe–Cu–K catalyst in a micro-fixed-bed based on the formate mechanism [88]. The optimal model shows that the slowest step in WGS reaction is desorption of gaseous carbon dioxide via formate intermediate species [89]. Table 8 summarizes the resulting mechanism. The parameters for CO2 formation appear in Table 9 [88]. Botes investigated the kinetic modeling of the WGS reaction in the iron-based low-temperature FTS. He found that WGS rate expressions based on the formate mechanism is an improved description of the WGS kinetic data. He selected three reaction schemes from the sets of elementary reactions presented in Table 8 (Models WGS-III5, WGS-V7 and WGS-IV6). Table 9 summarizes the rate expressions for these models with corresponding model parameters in Table 10.


13

Table 9 WGS reaction micro kinetic rate expressions. Model

WGS rate equation based on formate intermediate mechanism

High temperature WGS III5

2 RCO2 ¼ kw PCO PH2 O PCO2 PH2 =K p = 1 þ K 1 PCOþ K 3 PH2 O

Side balances

Ref.

s þ COs þ H2 Os

[77]

s þ COs þ OHs

[77]

s þ COs þ H2 Os

[77]

s þ COOHs

[77]

s þ H2 Os þ OHs

[77]

s þ COs þ OHs

[77]

s þ COOHs

[77]

s þ OHs

[77]

s þ CO2 s

[77]

s þ COs þ OHs

[85]

CHs þ HOs

[82]

CO þ s2

[86]

COs2 þ OHs2

[86]

COOHs

[86]

COs þ OHs

[86]

H2 þ 2s

[86]

COs þ H2 Os

[86]

CHOs þ HOs

[86]

H2 Os2 þ s2

[86]

COs þ H2 Os

[86]

s2 þ COs2 þ H2 Os2

[27]

s2 þ COs2 þ H2 Os2

[27]

1

WGS IV6

RCO2

WGS IV7

kw ¼ K 5 K 1 K 3 mmolgcat s1 bar2 2 ¼ kw PCO PH2 O =PH2 0:5 PCO2 PH2 0:5 =K p = 1 þ K 1 PCO þ K 3 PH2 O PH2 0:5 1 kw ¼ K 5 K 1 K 3 K 4 K s 0:5 mmolgcat s1 bar1:5

RCO2

WGS IV7

K ¼ K 4 K 3 =K s 0:5 2 0:5 ¼ kw PCO PH2 O =PH2 PCO2 PH2 0:5 =K p = 1 þ K 1 PCO þ K 3 PH2 O 1 1 0:5 1:5 kw ¼ K 5 K 1 K 3 K 4 K s mmolgcat s bar

RCO2 ¼ kw ðPCO PH2 O PCO2 PH2 =K p Þ=ð1 þ KPCO PH2 O =PH2 0:5 Þ2 1 kw ¼ K 1 K 3 K 4 K s 0:5 K 6 ðmmolgcat s1 bar1:5 Þ K ¼ K 1 K 4 K 3 K s 0:5 =K 6 2 RCO2 ¼ kw PCO PH2 O PCO2 PH2 =K p = 1 þ K 1 PCO þ K PH2 O =PH2 0:5 1 kw ¼ K 7 K 1 K 3 K 4 K 6 mmolgcat s1 bar2

WGS V7

K ¼ K 4 K 3 =K s 0:5 2 RCO2 ¼ kw PCO PH2 O PCO2 PH2 =K p = 1 þ K 1 PCO þ KPH2 O =PH2 0:5 1 kw ¼ K 7 K 1 K 3 K 4 K 6 mmolgcat s1 bar2

WGS V8

WGS V8

WGS VI4

RCO2

K ¼ K 4 K 3 =K s 0:5

2 RCO2 ¼ kw PCO PH2 O PCO2 PH2 0:5 =K p = 1 þ KPCO PH2 O 1 1 3 kw ¼ K 8 K 7 K 6 K 5 K 4 K 3 K 1 mmolgcat s bar K ¼ K7 K6 K5 K4 K3 K1 2 ¼ kw PH2 O =PH2 0:5 PCO2 =PH2 0:5 =K p PCO = 1 þ KPH2 O =PH2 0:5 1 kw ¼ K 3 K 4 K 2 0:5 mmolgcat s1 bar0:5 K ¼ K 2 K 3 0:5 2 RCO2 ¼ kw PCO PH2 O =PH2 PCO2=K p = 1 þ KPCO PH2 O =P H2

WGS VI6

0:5

1 1

2

kw ¼ K 6 K 1 K 2 K 3 K 4 K 5 mmolgcat s bar RCO2 ¼ kv PCO PH2 O =PH2 0:5 PCO 2 PH2 0:5 = 1 þ K v PCO PH2 O =PH2 0:5

WGS-VIII4

0:5

WGS- VII1 WGS-I1 WGS-II6 WGS-VIII4 WGS-IV6

WGS-IV5

WGS-VIII3

WGS-V7

WGS-II4

WGS-III5

K v ¼ K P K WGS;4 K WGS;5 kv ¼ kWGS;4 ¼ kWGS;4 =kWGS;4 5078:0045 5:8972089 þ 13:958689 104 T 27:592844 108 T 2 lnK P ¼ T pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi K w3 PH2 þ PH2 þ K w1 PCO P H2 O =PH2 RCO2 ¼ kw2 K w1 K w3 PCO PH2 O kw2 PH2 PCO2 = RCO2 RCO2 RCO2

kw2 ¼ kw2 K w1 K w3 =K P ¼ kv PCO PCO2 PH2 PH2 O 1=K P = 1 þ kv PCO PH2 =PH2 O 2 ¼ kw PCO PH2 O =PH2 0:5 PCO2 PH2 0:5 =K p = 1 þ K v PH2 O =PH2 0:5 ¼ kv PCO PH2 O =PH2 0:5 PCO 2 PH2 0:5 =K P = 1 þ K v PCO PH2 O =PH2 0:5 2 RCO2 ¼ kv PCO PH2 O =PH2 0:5 PCO2 ð1=Ke; WGSÞPH2 0:5 = 1 þ K v PH2 O =PH2 0:5 kv ¼ kIV2 K IV1 K IV3 K IV4 K IV5 mol= gcat :s:bar1:5 K v ¼ K IV4 =K IV5 bar0:5 2 RCO2 ¼ kv PCO PH2 O =PCO2 1=K e;WGS PH2 = 1 þ K v PH2 O PCO =PCO2 1:5 kv ¼ kIV2 K IV1 K IV3 K IV4 K IV5 mol= gcat :s:bar K v ¼ K IV1 K IV2 K IV3 K IV4 bar0:5 RCO2 ¼ kv PCO PH2 O =PCO2 1=K e;WGS PH2 PCO2 =ð1 þ K v PCO Þ2 2 kv ¼ kVIII2 K VIII1 mol= gcat :s:bar K v ¼ K VIII1 bar1 2 RCO2 ¼ kv PCO PH2 O 1=K e;WGS PH2 PCO2 = 1 þ K v PH2 O =PH2 0:5 kv ¼ kV2 K V1 K V5 mol= gcat :s:bar2 K v ¼ K V5 =K V4 bar0:5 2 RCO2 ¼ kv PH2 O 1=K e;WGS PH2 PCO2 =PH2 O = 1 þ K v PH2 kv ¼ kII4 ðmol=ðgcat :s:bar ÞÞ K v ¼ K II3 bar1 RCO2 ¼ kv ðPH2 O PCO ð1=K e;WGS ÞPH2 PCO2 Þ=ð1 þ K v ðPH2 O þ PCO ÞÞ2 kv ¼ kIII3 K III1 K III2 ðmol=ðgcat :s:bar2 ÞÞ K v ¼ K III3 K III2 ðbar1 Þ

Low temperature WGS I5 WGS II6

RCO2

2 RCO2 ¼ kw PCO PH2 O PCO2 PH2 =K p = 1 þ K 1 PCO þ K 3 PH2 O kw ¼ K 5 K 1 K 3 molkgcat 1 s1 MPa2 2 ¼ kw PCO PH2 O =PH2 0:5 PCO2 PH2 0:5 =K p = 1 þ K 1 PCO þ K 3 PH2 O 0:5 1 1 1:5 kw ¼ K 5 K 1 K 3 K 4 K s molkgcat s MPa

14


Table 9 (Continued) Model

WGS rate equation based on formate intermediate mechanism RCO2 ¼ kv PCO PH2 O =PH2 0:5 PCO 2 PH2 0:5 = 1 þ K v PCO PH2 O =PH2 0:5

WGS-VIII4

Side balances

Ref.

s þ COs þ OHs

[87]

s þ COs þ H2 Os

[55]

COs þ OHs

[55]

COs þ OHs

[55]

K v ¼ K P K WGS;4 K WGS;5 0:5 kv ¼ kWGS;4 ¼ kWGS;4 =kWGS;4 5078:0045 5:8972089 þ 13:958689 104 T 27:592844 108 T 2 lnK P ¼ T 2 RCO2 ¼ A PCO PH2 O PCO 2 PH2 ð1=K WGS Þ = 1 þ K CO PCO þ kH2 O PH2 O 2 RCO2 ¼ A PCO PH2 O PCO 2 PH2 ð1=K WGS Þ = 1 þ kOH PH2 O =PH2 0:5 þ kH2 O PH2 O 2 RCO2 ¼ A PCO PH2 O =PH2 0:5 PCO 2 PH2 0:5 ð1=K WGS Þ = 1 þ kOH PH2 O =PH2 0:5 þ kCO PCO

WGS-III5 WGS-V7 WGS-IV6

Botes found that the Model WGS-V7 predicts the reaction rate accurately to within about 20% of the measured values over the whole range of pressures [55]. Another mechanism for the water–gas shift reaction is direct oxidation [90]. It has been proposed that oxidation of adsorbed or gas-phase CO to CO2 [54,91–93], presented in Fig. 6, best describes FTS on iron catalysts. The oxygen intermediate forms from the dissociation of water or CO. Rethwisch and Dumesic [91] studied the WGS reaction on several supported and unsupported iron oxide and zinc oxide catalysts. They suggest that the WGS reaction

Kinetic parameter

High temperature WGS-III5 kw ¼ 0:25 K 1 ¼ 0:39 K 3 ¼ 3:54 WGS-IV6 kw ¼ 1:21 K 1 ¼ 0:59 K 3 ¼ 12:69 WGS-IV6 kw ¼ 0:34 K 1 ¼ 0:38 K 3 ¼ 2:18 WGS-IV7 kw ¼ 0:03 K ¼ 0:29 WGS-V7 kw ¼ 0:04 K 3 ¼ 0:001 K ¼ 3:85 WGS-V7 kw ¼ 1:61 K 1 ¼ 0:69 K ¼ 27:46 WGS-V8 kw ¼ 0:01 K ¼ 0:11 WGS-VI4 kw ¼ 0:13 K ¼ 1:52 WGS-VI6 kw ¼ 0:12 K ¼ 3:68 WGS-VIII4 K v ¼ 18:34 WGS- VII1 K w4;0 ¼ 7:10 105 Ev ¼ 81:96

WGS-VIII4

Dimension

Ref.

mmol gcat1 s1 bar2 bar 1 bar 1 mmol gcat1 s1 bar1.5 bar 1 bar 1 mmol gcat1 s1 bar1.5 bar 1 bar 1 mmol gcat1 s1 bar1.5 bar 1 mmol gcat1 s1 bar2 bar 1 bar 1 mmol gcat1 s1 bar2 bar 1 bar 1 mmol gcat1 s1 bar3 bar 1 mmol gcat1 s1 bar0.5 bar 1 mmol gcat1 s1 bar1 bar 1 MPa0:5 mol kg1 s1 bar1

[77]

1

[77]

[77]

[77] [77]

[77]

[77] [77] [36] [82]

kJ mol bar 2

[82] [82]

K w3 ¼ 2:21 104

bar

[82]

K v ¼ 2:76 102

bar

Macro-kinetics or global kinetics empirically describes WGS rate equations with no mechanistic implications.

rWGS ¼ AP CO

ð17Þ

The most obvious weakness of Eq. (17) is the fact that it does not account for the reversibility of the WGS reaction; hence the provision that it is only applicable when the reaction is still far from equilibrium. Nevertheless, since other researchers have also reported a good correlation between Eq. (17) and their own data [55,95], there appears to be some consistency in the empirical observation that the WGS reaction rate is directly proportional to the CO partial pressure. Eq. (17) can be expanded to account for the reversibility of the WGS reaction by assuming that the driving force for the WGS reaction is the difference between the actual CO partial pressure and its corresponding equilibrium value [55]. The reaction rate is presented in Table 11. Shen et al. [96] measured the WGS kinetics on the same commercial catalyst in a gas–solid Berty reactor. They fit their data

[77]

K w1 ¼ 2:65 102

0.5

Macro kinetics

High-temperature WGS in FTS Dry described WGS rate as a first-order in CO because the WGS reaction was still far from thermodynamic equilibrium [94]:

Table 10 Model parameters for micro kinetic rate expressions of WGS reaction. Model

over unsupported magnetite proceeds via a direct oxidation mechanism while all supported iron catalysts operate via a mechanism with formate species due to limited change of oxidation state of the iron cations [27].

[86]

Table 11 Macro-kinetic rate expressions for WGS reaction. Reaction rate

Ref.

High temperature PH PCO2

2 rWGS ¼ AðPCO PCO equilibrium Þ ¼ AðPCO K WGS P

rWGS ¼ kWGS ðPCO PH2 O

H2 O

Þ

PH2 PCO2 K WGS Þ

rWGS ¼ kWGS Pf ðyCO y 0

CO Þ PH PCO2

WGS-III5 WGS-V7 WGS-IV6

¼ 1:77 ¼ 2:10 ¼ 24:19 ¼ 1:13 ¼ 2:78 ¼ 12:27

¼ 1:13 103 kCO ¼ 0:5 kH2 O ¼ 3:4 kH2 O ¼ 1:1 kOH ¼ 6:3 kH2 O ¼ 0:7 kOH ¼ 21:9

1 1

2

mol kg s MPa MPa1 MPa1 mol kg1 s1 MPa1.5 MPa1 MPa1 bar0.5

[27]

bar1 bar1 bar1 – bar1 –

[55]

[27]

[167] [183,167] [183,167,190]

rWGS ¼ kWGS P f ðPCO K WGS PH O Þ 2

2

E

Low temperture WGS-I5 kw K1 K3 WGS-II6 kw K1 K3 WGS-VIII4 Kv

[55]

ri ¼ K i;0 expð RTa;i ÞPCO mi PH2 ni K i;0 = 4.799*108, m = 0.57, n = 0.7, Ea;i = 58,800 (J/mol)

[98]

rWGS ¼ kWGS :ðPCO PH2 O K 1 ðPCO2 PH2 ÞÞ=ðPCO þ K 2 PH2 O Þ2 kWGS = 0.0292 mol/kg.s, K 1 = 85.81, K 2 = 3.07

[99]

RWGS ¼

0:5 15:7expð45080=RTÞ PCO PH2 O =P0:5 H P CO2 P H =K WGS

2

1þ1:13103 PCO PH2 O =P0:5 H

[168]

2

2

[87]

[55]

5:8972089 þ 13:958689 K WGS ¼ 5078:0045 T 104 T 27:592844 108 T 2 Low temperature rWGS ¼ kWGS

[55]

ðPCO PH2 O PH2 ðPCO2 =K eq ÞÞ PCO þcPH2 O

K eq ¼ 0:0132expð4578=TÞ

[95,102,77,100]


CO

O CO2

CO2

O

or

CO2

15

different investigators. The water–gas shift reaction rate (Eqs. (18) and (19)) was proposed by Kuo et al. [102] and Zimmerman et al. [95]. For the cobalt-based catalyst, the rate equation was taken from Satterfield and Yates [78,100]: rWGS ¼ kWGS

ðPCO PH2 O PH2 ðPCO2 =K eq ÞÞ PCO þ cPH2 O

ð18Þ

CO K eq ¼ 0:0132expð4578=TÞ

or H2O

OH C

O

ð19Þ

The next section examines H2 perm-selective membrane reactors with an emphasis on WGS reaction rates in tubular FTS reactors. H2 perm-selective membrane reactor

CO H Fig. 6. Water–gas shift reaction via direct oxidation.

with the same equations and found activation energies for kWGS in Eq. (18) up to 88 kJ/mol [97]. Montazer-Rahmati et al. [98] obtained the intrinsic reaction rates for the FTS on a modified iron catalyst made in the Gas Research Department of RIPI. They presented a number of powerlaw rate equations based on their experiments. After that, they used these rate equations for a pilot plant design. The production rate of each product is a function of the partial pressures of carbon monoxide and hydrogen. Table 11 summarizes the reaction rate and kinetic parameters. Raje and Davis [99] presented the kinetic model for the FTS and WGS reactions on an iron catalyst in a slurry reactor (Table 11). Low-temperature WGS in FTS Inga et al., unlike others, takes into account the significance of WGS reaction in the gas-liquid mass transfer calculation and makes use of the “singular kinetic path” [100] concept proposed by Espinoza [101], which incorporates FT kinetic expressions from

Hydrogen perm-selective membrane systems efficiently separate hydrogen from a feed stream in a variety of industrial processes [103], including gasification, reforming, and petrochemical processes. Fig. 7 shows a generalized flow diagram of H2 production and separation from carbon-containing feed stocks. In configuration A, a membrane separates hydrogen from CO in a syngas feed stream produced by gasifying fossil or biomass fuels or by steam reforming of natural gas. In configurations B and C, additional syngas reacts with steam in a water–gas shift reactor to produce additional hydrogen [104]. Alternatively, in configuration B, the membrane can be combined with a WGS catalyst for improved CO conversion by removing produced hydrogen. In configuration C, the membrane operates after a WGS reactor as in applications of IGCC-CCS power plants and ammonia synthesis. Finally, in configuration D, the membrane removes hydrogen during dehydrogenation of an alkane feed stream produced in petrochemical processing, increasing the efficiency of olefin production. Palladium-based membrane reactor Palladium-based membranes enjoy a reputation as the most suitable membranes for hydrogen production [105]. Graham

Fig. 7. H2 production and separation from carbon containing feed stocks and use of membrane in various applications.

16


Fig. 9. Schematic of inert membrane catalytic packed bed reactor. Fig. 8. A mechanism of H2 permeation through Pd-based films.

reported the first investigation of hydrogen absorption and diffusion through palladium membranes in 1866 [106]. In the last few years, important studies related to hydrogen generation have been performed using palladium membrane reactors [107]. Pure Pd-membrane reactors undergo phase transition from a to b at temperatures and pressures below 300 C and 2 MPa, respectively, depending on the metal concentration. This transition leads to lattice strain and, consequently, after a few cycles [108], distorts the metal lattice because the lattice constant of the b-phase is 3% larger than that of the a-phase [109]. Palladium alloys, particularly silver alloys, reduce the critical temperature for this defect and increase hydrogen permeability. Many hydrogen generation systems deploy Pd–alloy membranes on a stainless-steel support [110]. The maximum permeability of hydrogen uses an alloy with 23 wt.% silver [111]. Hydrogen permeates across a thin palladiumbased membrane with a finite selectivity [105] due to the difference in hydrogen partial pressure between two sides of the membrane [105]. As depicted in Fig. 8, the permeation takes place in a multi-step process. First, molecular hydrogen absorbs on a palladium surface and then dissociates into atomic hydrogen. The hydrogen atoms then diffuse through palladium lattice while the electrons interact with the metal lattice on the adjacent side [112]. Finally, hydrogen atoms leave the lattice and recombine on the surface before they desorb as hydrogen molecules [113]. Nb, Ta, and V have higher-hydrogen permeability values than palladium and palladium alloys. Also, some low-cost metals such as Ni and Fe have significant permeability. Unfortunately, Ni and Fe are more resistant to hydrogen transport than Pd. Thin palladium coatings over these metals avoid the formation of oxides on the metallic surfaces and reduce hydrogen adsorption activation energy [114] and, consequently, increase hydrogen permeation flux [115]. Oertel et al. analyzed a reactor with a 100 mm thick Pd membrane [116] for the first time. At high process temperatures (700–800 C), the sizable membrane produced low hydrogen permeability. Dense Pd and Pd alloy membranes can separate hydrogen [117] and its isotopes from gas mixtures [118]. For example, a Pd–Ag alloy with 20–25 wt.% Ag is commercially used to purify hydrogen [119]. Palladium-based membranes have a long history of hydrogen extraction because of their high permeability and good surface properties and because palladium is 100% selective for hydrogen transport [120,121]. Membrane reactor types used for hydrogenation and dehydrogenation reactions Generally, hydrogenation reactors consist of a long vertical container with a height-to-width ratio greater than two. They are divided into numerous tall chambers to enhance fluid distribution through the solid catalyst [122], such as a palladium-based catalyst [123], a polymeric-based catalyst, etc [124]. Vertical baffles increase the effective height-to-width ratio and therefore the linear velocity of the flowing reactants by several fold compared to

a single pass reactor with the same overall space velocity. Also, the baffles increase the vessel strength and ease of loading, regeneration and dumping of the catalyst. Some chambers should have the inlet and outlet nozzles at the bottoms of both the first and last chamber to enhance flow control [122]. Hydrogenation reactors are classified into conventional and membrane reactors. In the last decade, researchers have investigated reactor with the following membrane types: palladium [125], polymeric catalytic [124], autothermal, zeolite [126], ceramic [127], carbon [128], inert [129], and dense [130]. Reactors include fixed bed, packed bed and fluidized bed design, each of which is discussed separately below. Packed bed membrane reactor Palladium-based, perm-selective membranes typically use a packed bed of catalyst (Fig. 9). Hydrogenation reaction takes place on the packed bed catalyst [131]. The generated hydrogen permeates through the membrane due to the considerable difference in hydrogen partial pressures between the reaction and the permeation sides [132]. The mass transfer driving force is created by three different mechanisms; 1. An inert sweep gas on the permeation side (e.g., nitrogen, helium, etc.) [112]. 2. A higher pressure in the retentate than the permeate channel (if necessary by evacuation of permeate) [112]. 3. A reactive sweep gas to consume the permeated hydrogen (e.g. oxygen, air, carbon monoxide, etc.) [112]. Fixed-bed membrane reactor Fixed-bed membrane reactors consist of large cylinders as well as jacketed or shell-and-tube reactors containing one or several hollow tubes. The feed enters the rector and reacts on the catalyst bed packed in the annulus. Recently, chemical processes based on hydrogenation or dehydrogenation reaction focus on membrane reactors such as the FMR [133]. Fluidized bed hydrogen perm-selective membrane reactors A membrane-assisted, fluidized-bed reactor (MAFBR) consists of two concentric tubes [134]. The inner and the outer tubes are the catalytic reaction and permeation sides, respectively [135]. The inner tube comprises a Pd-alloy membrane on a stainless steel support. Integrating a membrane within a fluidized bed reactor enhances the conversion and selectivity of both the membrane and fluidized-bed reactor [133]. Other advantages include: 1. Isothermal operation [136]. 2. Negligible pressure drop [136]. 3. Arrangement of membrane package with flexibility in membrane and heat transfer surface [133]. 4. Improved fluidization behavior owing to the compartmentalization and also reduction of average bubble size.


Relatively few investigations appear of the combination of fluidized bed operation with membrane concept despite its potential advantages. Adris et al. [137,138] proved the concept and usefulness of this combination and also investigated the hydrodynamic aspects of this special type of fluidized bed, which involves small internal spacing. In fluidized bed reactors, force and vigorous particle motion lead to severe mechanical damage. Internal tubes in fluidized beds can be considered for heat transfer purposes. The lifespan of these internal tubes can be compromised by solid particles. The application of vertical tubes in bubbling fluidized bed is more common than horizontal tubes due to their advantages for construction, bed hydrodynamics, and pressure drop [122]. However, more selective removal of the product can be achieved with horizontal membranes [139]. A multi staged fluidized bed configuration with a fresh sweep gas in each stage can be applied with a vertical membrane configuration. Most applications use a vertical tube bundle [140]. Alternatives for hydrogen production reactions The interaction between hydrogen and palladium modifies the metal lattice leading to membrane expansion (contraction) as a result of hydrogen desorption. A Pd-membrane reactor should be tightly sealed to the membrane module to guarantee the selective separation of hydrogen [141]. Recently Hashi et al. [142] focused on several metals such as Nb, V, Ta, Ni and Ti alloys to avoid or reduce the use of Pd. A great number of reactions can be carried out to generate hydrogen using metal membrane reactors.

NG NG feed/ f ed/ fe Refinery Refi f nery off-gas off ff-gas

Water–gas shift process (WGS) WGS reactions produce hydrogen from syngas for fuel cell applications, adjust hydrogen-to-carbon monoxide ratio of synthesis gas, and also produce hydrogen for ammonia synthesis [143]. Membrane reactors can improve H2 production from the equilibrium-limited WGS reaction [144]. In accordance with Eq. (1), by transporting hydrogen through the membrane, the reaction overall amount of hydrogen increases even though equilibrium does not change [145], which gives a higher overall conversion compared with equilibrium values. Membrane reactors separate and purify hydrogen isotopes in fusion research reactors [146]. Fig. 10 shows production and purification of hydrogen based on WGS membrane reactor unit. Steam reforming process Steam reforming produces about half of all hydrogen and therefore represents an important potential application of membrane technologies. Steam reforming and similar technologies such as partial oxidation, and auto-thermal reforming produce hydrogen from fossil and other fuels. Fig. 11 shows a membrane reactor for auto-thermal methanol reforming [147]. Membrane reactors reduce the reactor volume required in auto-thermal methanol reforming, but there is also a small reduction in the overall efficiency and more complicated process controls [148]. The startup time and load variability generally favor partial oxidation or auto-thermal reforming. Commonly, oxygen-blown systems are used to avoid diluting the hydrogen product with nitrogen. CO CO22 Product Product (Purity (Purity depends depends on on membrane membrane composition) composition)

Steam Steam

Recycled Recycled H H22

HDS unit

17

Reformer

WGS MR

NG NG fuel f el fu Fig. 10. Hydrogen production and purification based on WGS membrane reactor unit.

Fig. 11. Membrane reactor for Auto-thermal methanol reforming.

Pure Pure H H22

18


Fig. 12. Schematic of pure hydrogen production by hydrocarbon compounds steam reforming, (a) traditional scheme and (b) inorganic membrane reactor [114].

with C2 or higher fuels frequently suffers from carbon deposition and poisoning on the catalyst. Traditional MSR processes include three distinct steps: steam methane reformer, a water–gas shift reactor to increase H2 concentration, and hydrogen purification [156]. Purification is commonly performed with pressure swing absorption (PSA) to produce 99.99% hydrogen. As seen in Fig. 13, the membrane reactor can combine these three steps in a single reactor [156].

A membrane reformer can both produce and separate hydrogen in a single unit. The membrane reactor technology is well documented for dense, self-supported Pd–Ag membranes [149] as well as Pd layers deposited on different substrates [150]. Dense metallic Pd-membranes produce very high H2 perm-selectivity with respect to other dense metallic compounds [151], but limited Pd availability makes it very expensive. Porous membrane substrates such as alumina, silica oxide or titania are stable at high temperatures, chemically inert and have relatively high permeability and moderate cost [152]. On the other hand, they decrease reactivity and selectivity due to permeation through the membrane [153]. Practitioners and researchers have demonstrated the importance of membrane fabrication in hydrogen selectivity [154]. Fig. 12 illustrates a two-step water–gas-shift reactor followed by a separation/purification unit b. Methane-steam reforming over palladium currently produces more hydrogen than any other process. It operates stably even though hydrogen partial pressure decreases in the catalyst bed because of hydrogen removal [155]. Hydrocarbon-steam reforming

Dehydrogenation process Selective dehydrogenation of unsaturated alcohols produces specialty chemicals such as unsaturated aldehydes used in fragrances and pharmaceuticals. It is difficult to dehydrogenate hydroxyl groups selectively because the hydroxyl groups and carbon–carbon double bonds in unsaturated alcohols both tend to react to consume hydrogen [158]. Hydrogen permeable membrane reactors such as palladium membrane reactors may be more successful because the hydrogen is withdrawn from the reaction

Steam Steam

NG NG feed f ed fe

Pre-treatment

Steam reforming

WGS

Hydrogen purification (PSA)

NG NG fuel f el fu

Compression

CO, CO, CO CO22,, H H22,CH ,CH44,, N N22,, H H22O O

Steam Steam

NG NG feed f ed fe

NG NG fuel f el fu

Pre-treatment

Reforming and membrane purification

Compression

CO, CO, CO CO22,, H H22,CH ,CH44,, N N22,, H H22O O Fig. 13. Process scheme for MSR and MR.

Pure ure H H22

Pure Pure H H22


19

Table 12 Some theoretical and experimental investigations on Pd-alloy membrane assisted in different types of reactor configuration. Membrane material

Thickness (mm)

Reactor type

Sweep gas

Feed

Reaction

Main Product

Temperature (K)

Ref.

Pd–Ag Pd–Ag Pd–Ag Pd–Ag Pd Pd–Ag Pd–Ag Pd–Ag Pd–Ag Pd Pd–Ag Pd–Ag Pd Pd–Ag Pd–Ag Pd–Ag Pd–Ag Pd–Ru Pd–Ru Pd–Ag Pd–Ag Pd–Ag Pd–Ag Pd–Ag Pd Pd–Ag Pd–Ag Pd–Ag Pd Pd Pd Pd–Ag Pd–Ag Pd–Ag Pd–Ag Pd Pd–Ag Pd–Ag Pd–Cu Pd

– 50 6 10 13 50 5 70 60 1 60 50 – 50 50, 60 50 20 254 254 Variable 10 60 10 6 1,600 200 Variable 50–60 – 50–150 – 6 70 50 50 1 50 60 125 200

FMR MR FR MRL MR PMR CMR MR MR MR MR MR CMR CMR MR MTMR MR MR MR MR CMR MR FBMR FMR MR MR MR MTMR MR MR MR CMR MR Flat-MR MR MR MR MR MTMR MR

Ar – Ar – Ar N2 – – – – N2 Ar – N2 N2 Steam N2 N2 N2

Dehydrogenation Gas detritiation Dehydrogenation Methanol synthesis SR WGS Ammonia decomposition WGS WGS Hydroxylation OSR CO2 reforming of methane MSR ESR ESR ESR OSR Dehydrogenation Dehydrogenation MeOH synthesis Ammonia decomposition WGS DME synthesis Dehydrogenation DR Dehydrogenation MSR ESR MSR MeOH synthesis Thermal decomposition Dehydrogenation WGS Methane decomposition WGS Direct hydroxylation MeOH SR WGS WGS Dehydrogenation

Benzene Hydrogen Benzene Methanol Hydrogen Hydrogen NOx H2 H2 Phenol H2 H2 H2 H2 H2 H2 H2 H2 H2 MeOH H2 H2 DME Ethylene H2 Hexanal H2 H2 H2 MeOH H2 Ethylene H2 Higher alkanes H2 Phenol H2 H2 H2 H2

503 573, 673 503 Variable 773 473–573 – 500–700 553–593 Variable 673,723 520–570 Variable 673 673–723 633 593–723 773 823 502 – 533–723 Variable Variable 783 443–503 750 963–1163 773 503 973–1273 660 604–623 318,373 598–603 423 623–823 493–593 1173 573

[191] [117] [192] [193] [194] [124] [195] [126] [196] [197] [198] [199] [149] [150] [200] [201] [202] [203] [203] [204] [205] [206] [135] [207] [208] [162] [209] [210] [211] [212] [213] [214] [215] [216] [116] [217] [218] [133] [219] [220]

Pd Pd–Ag Pd Pd Pd–Ag Pd–Ag Pd–Cu Pd–Ag Pd Pd Pd–Ag Pd–Au Pd Pd–Ag Pd Pd Pd–Ag

– 50 80–315 1 15 10 50 50 4.5 – 10 0.5–80 10 800 25 10 5

MR MR CMR PSMR SSMR CMR MMR MR FBMR CMR RFTMR MR MR MR MR MR HFMR

Cyclohexane Helium, Tritiated water Cyclohexane Synthesis gas Dodecane CO, H2O, CO2, H2 Synthesis purge gas CO, H2O CO, H2O Benzene EtOH CO2,CH4 Natural gas Bio-EtOH, H2O EtOH, H2O EtOH, H2O EtOH, H2O Iso-C4H10 C3H8 Syn Gas Urea wastewater CO, H2O Synthesis gas C2H6 CH4 cis-3-hexen-1-ol CH4 EtOH, H2O Natural gas Synthesis gas Hydrogen Sulfide EtOH CO, H2O Methane CO,H2O Aromatic compounds H2O, CH3OH CO,H2O CO,H2O Cyclohexane, Methylcyclohexane Alkane EtOH Synthesis gas Methane Methane, H2O Naphtha Benzene EtOH EtOH Ammonia Naphtha CO, H2O, CO2, H2 Ethylbenzene Synthesis gas EtOH CO,H2O CO,H2O, Ar

Dehydrogenation ESR MeOH synthesis MSR MSR Naphtha reforming Direct hydroxylation Partial oxidation EAR Ammonia decomposition Naphtha reforming WGS Dehydrogenation MeOH synthesis ESR WGS WGS

H2 H2 H2 H2 H2 Gasoline Phenol H2 H2 H2 Gasoline H2 Styrene MeOH H2 H2 H2

673 873 443–500 873 943–1053 777 423–523 723 973 823 777 723 833 401 673 573 723

[143] [221] [222] [223] [224] [118] [225] [226] [227] [228] [229] [230] [231] [232] [200] [233] [234]

– – Syn gas – – – – N2 Steam – – N2 N2 N2 N2 He N2 – Ar –

– N2 N2 – – N2 N2 – – – – N2 – – – Ar

side to the permeation side [159]. Palladium membrane reactors have the capability to promote dehydrogenation of cyclohexane to form benzene [160,161]. Hydrogen permeable membrane reactors are useful in the selective conversion of unsaturated alcohols to unsaturated aldehydes because of hydrogen of removal [162]. Table 12 presents some theoretical and experimental investigation on Pd-alloy membrane assisted in different types of reactor configuration. Pd-membrane development Several opportunities exist to improve Pd-based membrane reactor efficiency. Pd-based membrane reactors commonly deploy

self-supporting metal foils with thickness ranges of 25–100 mm. However, these membranes are expensive and have a low hydrogen flux. Decreasing the thickness increases chemical performance but decreases mechanical strength. Metallic membranes should be deposited on strong supports to achieve both high selectivity and good mechanical strength. A metal alloy is commonly used on the tubular structure via sequential electrolysis plating followed by alloying. This method is the most cost effective technique and has the potential to be industrialized [163]. It is possible to fabricate membrane layers such as Pd or (Pd-30%Ag) with a thickness of 3–5 m on commercially available ceramic supports by optimizing the electrolysis method. Typical tube lengths and outer diameter are 0.6–0.85 m and 14 mm. In the study

20


Table 13 Hydrogen separation efficiency of Pd membranes. Membrane

Thickness (mm)

Temperature ( C)

Driving force (MPa)

H2 flux (mol/m2 s)

Ref.

Pd/HF Pd/MPSS Pd/MPSS Pd/PG Pd–Ag/PG Pd/Al2O3 Pd–Ag/MPSS Pd–Cu/Al2O3 Pd–Cu/Al2O3 Pd–Ag Pd/MPSS Pd–Ag/MPSS Pd–CeO2/MPSS

3–4 10 19–20 13 21.6 0.5–1 15 3.5 1.5 50 6 4 13

430 480 500 500 400 350–4,500 500 350 350 500 550 500 500

0.1 0.1 0.101 0.202 0.202 0.1 0.202 0.1 0.1 0.1 0.1 0.1 0.2

0.136 0.089 0.0150–0.030 0.189 0.067 0.05–0.1 0.103 0.056 0.499 0.010 0.300 0.280 0.275

[235] [236] [237] [238] [238] [239] [240] [241] [243] [242] [243] [243] [244]

of Paglieri and Way [164], membranes were applied for single gas permeability tests at different temperatures and also for separation of hydrogen from reformate gas, using a bench-scale test system that can operate up to 500 C and 65 bar feed pressure with a membrane area of about 50 cm2. The results showed that, after initial activation, considerable hydrogen permeances of 5–100 m3/ m2h bar0.5 were obtained. Table 13 shows hydrogen separation efficiency of Pd membranes [115]. Tubular reactor models for WGS in FTS The tubular reactor for FTS consists of a multi-tube reactor where the synthesis gas is fed at the top of the reactor and passes through a catalyst bed inside the tubes. The synthesis gas passes through the reactor in a plug-flow regime. Water passes through the shell side of the multi-tube reactor to remove the heat released by the FTS reaction and to control the reactor temperature, forming a nearly isothermal reactor in spite of the high heat of reaction [120,165]. A schematic tubular reactor is shown in Fig. 14. Hightemperature reaction rates of FTS differ from low-temperature rates; therefore modeling of FT reactor at high and low temperature differs. Mazzone et al. developed a mathematical model of a tubular reactor for syngas and carbon monoxide polymerization. They used the kinetic model for the FTS reaction on an iron catalyst and for the WGS reaction, which are given by Raje and Davis [99]. They concluded that the production of gasoline and diesel using this iron-based catalyst is not economically viable in this kind of reactor since their yields are lower than 5% in mass weight, though

Fig. 14. Schematic diagram of a conventional tubular reactor.

while some commercial catalysts claim yields higher than 15% in mass weight [165]. Marvast et al. [166] modeled a water-cooled fixed bed FT reactor packed with Fe-HZSM5 catalyst in two dimensions (radial and axial) using the intrinsic reaction rates previously developed by Montazer-Rahmati et al. [98]. They carried out a parametric sensitivity analysis and show that increasing the tube diameter has a minor effect on C5+ production and increasing the shell temperature is not recommended with respect to the C5+ and CO2 production rates [166]. Increasing the shell temperature decreases C5+ production and CO2 formation. Moreover, they found that increasing the H2/CO ratio increases overall yield of C5+ products [166]. CO2 production is severe in conventional reactors (tubular fixed bed reactors). New configurations are proposed for decreasing CO2 production and increasing gasoline production. Forghani et al. proposed a novel reactor configuration with a hydrogen perm-selective membrane for FTS. In this configuration, the synthesis gas enters the tube side and flows in co-current mode with a reacting gas mixture that enters the shell side of the reactor. In this way, the synthesis gas warms through the heat of reaction produced in the reaction side. Hydrogen can penetrate from the feed synthesis gas side into the reaction side as a result of a hydrogen partial pressure difference. The outlet synthesis gas from the tube side recycles to the shells and the chemical reaction is initiated in the catalytic bed. Therefore, the reacting gas on shell side cools simultaneously as the gas on tube side warms. This membrane Fischer–Tropsch reactor improves the selectivity of hydrogenation with hydrogen passing through a membrane and increases production of high-octane gasoline from synthesis gas on bi-functional Fe-HZSM5 catalyst [121]. They modeled this configuration using kinetic rates obtained by Montazer-Rahmati et al. [98]. They showed that the membrane reactor operates favorably compared to a conventional reactor, with decreased carbon dioxide and methane production and increased gasoline production [121]. A weakness of the tubular fixed-bed reactor is poor heat transfer and exothermic reactions on the tube side do not cool well. Better coolants on the shell side improve a tubular fixed bed reactor [167]. Jayhooni et al. proposed a novel configuration for a cooling system of an FTS reactor. In this new configuration, a nanofluid cools the inside shell of a fixed bed reactor. Heat transfer from tubes improves in these nano-fluid reactors (NFR) because they use aqueous fluids containing alumina nanoparticles. They used the kinetic model Montazer-Rahmati et al. for FTS and the kinetic model and Wang et al. [168] for WGS. They showed that the ratio of CO2 yield to the gasoline yield decreases in comparison with conventional reactors [169]. Rahimpour et al. proposed a one-dimensional heterogeneous model to analyze the performance of a fixed-bed reactor combined with a membrane assisted fluidized-bed reactor. In this new concept, the synthesis gas converts to FT products in two


Pd-Ag membrane tube

Steam reforming

21

Steam drum

Feed Syngas

Product

Reaction gas Fixed-bed reactor

Fluidized-bed membrane reactor

Fig. 15. Schematic diagram of fluidized-bed membrane followed by fixed-bed reactors.

catalyticreactors. The first reactor is a water-cooled fixed bed while the second reactor is a gas-cooled fluidized bed [110] (Fig. 15). They modeled this dual configuration with the kinetic model of MontazerRahmati et al. [98]. Results show improved gasoline yield, a decrease in CO2 formation and an improved temperature profile [110]. Rahimpour et al. investigated a novel configuration of FTS reactors in which a fixed-bed water perm-selective membrane reactor precedes a fluidized-bed hydrogen perm-selective membrane reactor (Fig. 16). They modeled this configuration with the Montazer-Rahmati et al. [98] kinetics. The water permeation through the H-SOD membrane layer leads to better H2O removal from the reaction side of the first reactor (the water-cooled reactor). It accelerates the reverse WGS reaction of CO2 with H2

towards CO, which reacts in a subsequent step towards long-chain hydrocarbons. Moreover, hydrogen permeates through the Pd-Ag membrane layer to the reaction side of the second reactor (the gascooled reactor). These two factors increase the gasoline production rate [170]. Rahimpour et al. proposed a thermally-coupled reactor containing the FTS reaction in the exothermic side and dehydrogenation of cyclohexane on the endothermic side with a hydrogen perm-selective membrane as the shell of the reactor to separate the hydrogen product from the dehydrogenation process. Permeated hydrogen enters another section called the permeation side where argon seeps it. They used a one-dimensional model for optimization of a reactor consisting of the FTS reaction coupled

N2 & H2O Steam reformer

Steam drum Synthesis gas Pd-Ag membrane

H-SOD membrane

Product

Reaction gas N2 Fluidized-bed membrane reactor

Fixed-bed membrane reactor

Fig. 16. Schematic diagram of fixed-bed membrane and fluidized-bed membrane reactors.

22


GTL product

Nitrogen

Hydrogen recycle

Gas decalin

Exothermic side H-SOD membrane

Synthesis gas

Pd-Ag membrane Recycle endothermic Naphthalene side

Catalyst

Hydrogen Stripper Hydrogen

Pure Hydrogen

Water & Nitrogen

Hydrogen & Nitrogen

Fig. 17. Schematic process diagram of the thermally coupled membrane reactor.

with dehydrogenation of cyclohexane. They used the kinetic model of Montazer-Rahmati et al. [98]. They optimized operating conditions of FTS for maximizing reactant conversion and hydrogen recovery and minimizing CO2 and CH4 yield. Modeling results show that this configuration can be a compelling way to boost hydrogen production and ensure the other desired results. However, an investigation in relation to the environmental aspects, commercial viability and economic feasibility of the proposed configuration is necessary to consider commercialization of the process [171]. Rahimpour et al. used a dehydrogenation reaction on the endothermic side of a coupled reactor containing the FTS reaction on the exothermic side [172]. Moreover, they used water and hydrogen perm-selective membrane layers [173] in this configuration (Fig. 17). Models show that CO2 yield in this configuration decreases for two reasons. Firstly, the in situ water removal from the exothermic side shifts the WGS reaction to the reactants side and CO2 is consumed and secondly, the high activation energy of the WGS reaction decreases at low temperatures [173]. Table 14 summarizes information about FTS reaction in above investigations. The CO2 and C5+ yields in this Table are for CO2 as a product of WGS and C5+ is a main product of FTS.

Future work Numerous studies on catalytic membrane reactors cited in the literature provide useful information [143]. Since the membrane reactor technology is not yet well established, there is no comprehensive information on the cost of this technology. Investigations should focus on developing alternative membranes with higher permeability and selectivity during a long period of time at more severe operating conditions. To improve membrane technology, some essential investigations are suggested as follows: 1. Fabrication of a thin membrane with high flux and suitable mechanical and chemical stability in all operating conditions [156]. 2. Investigation of hydrogen permeability of the membrane in the presence of potential coal gas contaminants such as HCl, NH3 and trace metals [174]. 3. A more detailed economic analysis of the membrane reactors. 4. Reduction of required membrane area and pressure vessel size by increasing the efficiency of membrane modules. 5. Optimal use of membrane reactors for minimization of catalyst and hardware cost.

Table 14 Results of mathematical modeling of tubular reactor FTS with different kinetic model of WGS. Kinetic model rWGS ¼ kWGS ðPCO PH2 O ðPCO2 PH2 ÞÞ=ðPCO þ K 2 PH2 O Þ E ri ¼ K i;0 exp RTa;i PCO mi PH2 ni E ri ¼ K i;0 exp RTa;i PCO mi PH2 ni

2

RWGS ¼

Specification of reactor

CO2 yield (%)

C5+ yield (%)

Ref.

Fe Fe-HZSM5

Conventional reactor

–

27

[165]

Conventional reactor

47

6.7

[166]

Fe-HZSM5 Fe-HZSM5

Membrane reactor

40

7.5

[121]

Conventional reactor with nano-fluid coolant

45

8.7

[169]

Dual type reactor (fixed bed and fluidized bed membrane reactor) Dual type membrane reactor

34

10.2

[110]

36

10.5

[170]

Coupling reactor

36

9.8

[171]

Coupling dual membrane reactor

42

8

[173]

0:5 15:7expð45080=RTÞ PCO PH2 O =P0:5 H P CO2 P H =K WGS

ri ¼ K i;0 ri ¼ K i;0 ri ¼ K i;0 RWGS ¼

Catalyst

exp

2

1þ1:13103 PCO PH2 O =P0:5 H Ea;i RT

PCO mi PH2 ni

2

2

Fe-HZSM5

E exp RTa;i PCO mi PH2 ni E exp RTa;i PCO mi PH2 ni

15:7expð45080=RTÞ

Fe-HZSM5

PCO PH2 O =P0:5 H2 P CO2

1þ1:1310

3

PCO PH2 O =P0:5 H2

P0:5 H2 =K WGS

Fe-HZSM5 Fe-HZSM5


6. Maximization of hydrogen recovery by a membrane with high separation factor. 7. Development of a process technology to separate hydrogen from a mixed gas feed stream for carbon capture in coal gasification facilities for power and chemicals production. 8. Energy penalty minimization for CO2 capture via process integration. 9. Use of high stability palladium alloys to tolerate severe process conditions, abrupt startups and shutdowns, and contaminants in feed streams. 10. Investigation on a cheaper and more stable alternative membrane material instead of palladium [156]. 11. Improvement the stability, reliability and lifetime of the membrane tubes at normal operating conditions [156]. Conclusion This paper reviews the reactions, kinetics and reactor design of WGS and FTS reactors. It also includes integration of reaction and separation in catalytic membrane reactors for the production of ultrapure hydrogen from hydrogen-containing gas mixtures using palladium-based membranes. The WGS reaction is a main methodof hydrogen production and is essential in many processes such as FTS. Iron-based catalysts provide both FTS and WGS reaction activities. The kinetic expressions of reactions are important for reactor design. The discussion reviews catalysts employed for both high- and low- temperature WGS and WGS-FTS reactions and the micro and macro kinetic approaches. Comprehensive lists of shift reaction models appear in Tables 6 and 7 for WGS and Tables 9 and 11 for WGS-FTS reactions, together with their respective parameters. The WGS rate can be calculated based on the data presented in the tables, which can be used as the references. Most authors employ a power-law kinetic expression, although different ranges of rates are obtained from experimental data. These kinetic expression assist reactor design. The last part of this investigation discusses mathematical models of tubular reactors using WGS reaction rate models. Since the main products of FTS are hydrocarbons (C5+), H2O, and CO2, the yields of CO2 and C5+ for all models are compared and comparison results are presented in Table 14. Conventional tubular reactors produce high CO2 and low C5+. Results show that, membrane reactor and coupling of endothermic with exothermic reactions lead to increasing C5+ and decreasing CO2 production. Palladium alloys, particularly with silver, should enhance hydrogen permeability and help manage costs. Pd–alloy membranes on stainless steel outperform other non-precious metals for hydrogen removal with high permeability and 100% selectivity for hydrogen transport. Many investigations during the last decade focus on the membrane reactor. The product yield and selectivity enhancement may distinguish membrane reactors from conventional rectors. This review emphasizes various hydrogen generation reactions and different types of Pd-based membrane configuration. Acknowledgements Our most sincere gratitude goes to Professor. Dr Mohammad Reza Rahimpour (Chemical Engineering Department, Shiraz University) for his fruitful advice during the manuscript preparation and Centre for Process Systems Engineering and Sustainability Pázmány Péter Catholic University in Budapest, Hungary for supporting the corresponding author’s research. References cek, J.J. Klemeš, P.S. Varbanov, Z. Kravanja, Clean Technol. Environ. Policy [1] L. Cu9 17 (2015) 2125.

23

[2] J.S. Cha, S.H. Park, S.-C. Jung, C. Ryu, J.-K. Jeon, M.-C. Shin, Y.-K. Park, J. Ind. Eng. Chem. 40 (2016) 1. [3] V.G. Dovì, F. Friedler, D. Huisingh, J.J. Klemeš, J. Clean. Prod. 17 (2009) 889. [4] A. Nemet, P.S. Varbanov, J.J. Klemeš, Clean Technol. Environ. Policy 18 (2016) 2029. [5] M.K. Nikoo, S. Saeidi, A. Lohi, Clean Technol. Environ. Policy 17 (2015) 2267. [6] S. Dutta, J. Ind. Eng. Chem. 20 (2014) 1148. [7] S.E. Hosseini, M.A. Wahid, Renew. Sustain. Energy Rev. 57 (2016) 850. [8] I.K. Kapdan, F. Kargi, Enzyme Microb. Technol. 38 (2006) 569. [9] F. Migliardini, O. Veneri, P. Corbo, J. Ind. Eng. Chem. 17 (2011) 633. [10] L. Singh, Z.A. Wahid, J. Ind. Eng. Chem. 21 (2015) 70. [11] M. Abashar, J. Ind. Eng. Chem. 21 (2015) 129. [12] Y.H. Jia, J.H. Ryu, C.H. Kim, W.K. Lee, T.V.T. Tran, H.L. Lee, R.H. Zhang, D.H. Ahn, J. Ind. Eng. Chem. 18 (2012) 715. [13] R. Sahu, B.J. Song, J.S. Im, Y.-P. Jeon, C.W. Lee, J. Ind. Eng. Chem. 27 (2015) 12. [14] N. Muhammad, Y.A. Elsheikh, M.I.A. Mutalib, A.A. Bazmi, R.A. Khan, H. Khan, S. Rafiq, Z. Man, J. Ind. Eng. Chem. 21 (2015) 1. [15] T. Abbasi, S. Abbasi, Renew. Sustain. Energy Rev. 15 (2011) 3034. [16] R.B. Gupta, Hydrogen Fuel: Production, Transport, and Storage, CRC Press, 2008. [17] S. Lee, J.-S. Kang, K.T. Leung, W. Lee, D. Kim, S. Han, W. Yoo, H.J. Yoon, K. Nam, Y. Sohn, J. Ind. Eng. Chem. (2016). [18] R. Lu, D. Mao, J. Yu, Q. Guo, J. Ind. Eng. Chem. 25 (2015) 338. [19] P. Parthasarathy, K.N. Sheeba, Int. J. Energy Res. 39 (2015) 147. [20] K. Pandu, S. Joseph, Int. J. Adv. Eng. Technol. 2 (2012) 342. lu, Int. J. Hydrogen Energy 26 (2001) 13. [21] D. Das, T.N. Vezirog [22] A.V. Bridgewater, Therm. Sci. 8 (2004) 21. [23] K. Meier, Int. J. Energy Environ. Eng. 5 (2014) 1. [24] B. Baghchehsaraee, G. Nakhla, D. Karamanev, A. Margaritis, Int. J. Hydrogen Energy 35 (2010) 5021. [25] C. Rhodes, G.J. Hutchings, A.M. Ward, Catal. Today 23 (1995) 43. [26] M.V. Twigg, Catalyst Handbook, Wolfe Publishing Ltd, 1989. [27] G.P. van der Laan, A.A.C.M. Beenackers, Appl. Catal. A: Gen. 193 (2000) 39. [28] J. Pasel, R.C. Samsun, D. Schmitt, R. Peters, D. Stolten, J. Power Sources 152 (2005) 189. [29] A. Nakhaei Pour, M.R. Housaindokht, S.M.K. Shahri, E.G. Babakhani, M. Irani, J. Ind. Eng. Chem. 17 (2011) 596. [30] S. Saeidi, M.T. Amiri, N.A.S. Amin, M.R. Rahimpour, Int. J. Chem. React. Eng. 12 (2014) 1. [31] D.S. Newsome, Catal. Rev. 21 (1980) 275. [32] C. Rhodes, B. Peter Williams, F. King, G.J. Hutchings, Catal. Commun. 3 (2002) 381. [33] M.L. Byron, R.J. Smith, M.S. Shan, Int. J. Chem. React. Eng. 8 (2010) 1. [34] C. Callaghan, I. Fishtik, R. Datta, M. Carpenter, M. Chmielewski, A. Lugo, Surf. Sci. 541 (2003) 21. [35] D.C. Grenoble, M.M. Estadt, D.F. Ollis, J. Catal. 67 (1981) 90. [36] R. Hakkarainen, T. Salmi, R.L. Keiski, Appl. Catal. A: Gen. 99 (1993) 195. [37] M. Sarkari, F. Fazlollahi, H. Atashi, Int. J. Chem. React. Eng. 10 (2012) 1. [38] E. Ruiz, D. Cillero, P.J. Martínez, A. Morales, G.S. Vicente, G. de Diego, J.M. Sánchez, J. CO2 Util. 8 (2014) 1. [39] T.J. Okeson, K. Keyvanloo, J.S. Lawson, M.D. Argyle, W.C. Hecker, Catal. Today 261 (2016) 67. [40] C. Rhodes, G.J. Hutchings, Phys. Chem.Chem. Phys. 5 (2003) 2719. [41] S. Saeidi, M.K. Nikoo, A. Mirvakili, S. Bahrani, N.A.S. Amin, M.R. Rahimpour, Rev. Chem. Eng. 31 (2015) 209. [42] H.F. Rase, Chemical Reactor Design for Process Plants: Case Studies and Design Data, John Wiley and Sons, USA, 1977. [43] M.V. Twigg, M.S. Spencer, Appl. Catal. A: Gen. 212 (2001) 161. [44] Y. Tanaka, T. Utaka, R. Kikuchi, K. Sasaki, K. Eguchi, Appl. Catal. A: Gen. 242 (2003) 287. [45] H. Kušar, S. Ho9 cevar, J. Levec, Appl. Catal. B: Environ. 63 (2006) 194. [46] D. Yeragi, N. Pradhan, A. Dalai, Catal. Lett. 112 (2006) 139. [47] T. Van Herwijnen, W. De Jong, J. Catal. 34 (1974) 209. [48] N. Amadeo, M. Laborde, Int. J. Hydrogen Energy 20 (1995) 949. [49] Y. Choi, H.G. Stenger, J. Power Sources 142 (2005) 81. [50] G. Jacobs, B.H. Davis, Appl. Catal. A: Gen. 333 (2007) 192. [51] C.V. Ovesen, P. Stoltze, J.K. Nørskov, C.T. Campbell, J. Catal. 134 (1992) 445. [52] G. Wang, L. Jiang, Y. Zhou, Z. Cai, Y. Pan, X. Zhao, Y. Li, Y. Sun, B. Zhong, X. Pang, W. Huang, K. Xie, J. Mol. Struct.: THEOCHEM 634 (2003) 23. [53] B. Arstad, H. Venvik, H. Klette, J. Walmsley, W. Tucho, R. Holmestad, A. Holmen, R. Bredesen, Catal. Today 118 (2006) 63. [54] C.V. Ovesen, B.S. Clausen, B.S. Hammershøi, G. Steffensen, T. Askgaard, I. Chorkendorff, J.K. Nørskov, P.B. Rasmussen, P. Stoltze, P. Taylor, J. Catal. 158 (1996) 170. [55] F.G. Botes, Appl. Catal. A: Gen. 328 (2007) 237. [56] K.C. Waugh, Catal. Today 53 (1999) 161. [57] J. Sun, J. DesJardins, J. Buglass, K. Liu, Int. J. Hydrogen Energy 30 (2005) 1259. [58] T. Salmi, R. Hakkarainen, Appl. Catal. 49 (1989) 285. [59] R. Hakkarainen, T. Salmi, R.L. Keiski, Catal. Today 20 (1994) 395–408. [60] R.L. Keiski, T. Salmi, P. Niemistö, J. Ainassaari, V.J. Pohjola, Appl. Catal. A: Gen. 137 (1996) 349. [61] I. Fishtik, R. Datta, Surf. Sci. 512 (2002) 229. [62] A.A. Gokhale, J.A. Dumesic, M. Mavrikakis, J. Am. Chem. Soc. 130 (2008) 1402. [63] N.Z.M.J.H. Mao, Pan Guo Xiang, Xu Qian, Acta Phys. Chim. Sin. 24 (2008) 2059. [64] L.C. Grabow, A.A. Gokhale, S.T. Evans, J.A. Dumesic, M. Mavrikakis, J. Phys. Chem. C 112 (2008) 4608.

24


[65] Q.-L. Tang, Z.-X. Chen, X. He, Surf. Sci. 603 (2009) 2138. [66] S.-J. Kim, I.-S. Byun, H.-Y. Han, H.-L. Ju, S.H. Lee, J.-G. Choi, Appl. Catal. A: Gen. 234 (2002) 35. [67] D. Wolf, M. Barre-Chassonnery, M. Höhenberger, A. Van Veen, M. Baerns, Catal. Today 40 (1998) 147. [68] R. Bove, P. Lunghi, N.M. Sammes, Int. J. Hydrogen Energy 30 (2005) 189. [69] W.F. Podolski, Y.G. Kim, Ind. 13 (1974) 415. [70] S.S.E.H. Elnashaie, S.S. Elshishini, Modeling, Simulation and Optimization of Industrial Fixed Bed Catalytic Reactors, OPA(Amsterdam) B.V, Gordon and Breach Science Publishers, 1993. [71] C.P.P. Singh, D.N. Saraf, Ind. Eng. Chem. Process Design Dev. 19 (1980) 393. [72] T.A. Adams II, P.I. Barton, Int. J. Hydrogen Energy 34 (2009) 8877. [73] H. Zhao, Y. Hu, J. Li, J. Mol. Cat. A: Chem. 149 (1999) 141. [74] D.F.B. Lima, F.A. Zanella, M.K. Lenzi, P.M. Ndiaye, Modeling and Simulation of Water Gas Shift Reactor: An Industrial Case, INTECH Open Access Publisher, 2012. [75] M. Levent, Int. J. Hydrogen Energy 26 (2001) 551. [76] C.H. Bartholomew, in: L. Guczi (Ed.), Chapter 5 Recent Developments in Fischer-Tropsch Catalysis, Elsevier, 1991, pp. 158. [77] A. Nakhaei Pour, M.R. Housaindokht, S.F. Tayyari, J. Zarkesh, S.M.K. Shahri, Chem. Eng. Res. Des. 89 (2011) 262. [78] I.C. Yates, C.N. Satterfield, Energy Fuels. 5 (1991) 168. [79] F. Fazlollahi, M. Sarkari, H. Gharebaghi, H. Atashi, M.M. Zarei, A.A. Mirzaei, W. C. Hecker, Chin. J. Chem. Eng. 21 (2013) 507. [80] M. Sarkari, F. Fazlollahi, H. Ajamein, H. Atashi, W.C. Hecker, L.L. Baxter, Fuel Process. Technol. 127 (2014) 163. [81] N.E. Amadeo, M.A. Laborde, Int. J. Hydrogen Energy 20 (1995) 949. [82] B.-T. Teng, J. Chang, J. Yang, G. Wang, C.-H. Zhang, Y.-Y. Xu, H.-W. Xiang, Y.-W. Li, Fuel. 84 (2005) 917. [83] A. Nakhaei Pour, M.R. Housaindokht, S.F. Tayyari, J. Zarkesh, J. Nat. Gas Chem. 19 (2010) 362. [84] D.B. Bukur, B. Todic, N. Elbashir, Catal. Today 275 (2016) 66. [85] J. Chang, L. Bai, B. Teng, R. Zhang, J. Yang, Y. Xu, H. Xiang, Y. Li, Chem. Eng. Sci. 62 (2007) 4983. [86] J. Yang, Y. Liu, J. Chang, Y.-N. Wang, L. Bai, Y.-Y. Xu, H.-W. Xiang, Y.-W. Li, B. Zhong, Ind. Eng. Chem. Res. 42 (2003) 5066. [87] E.S. Lox, G.F. Froment, Ind. Eng. Chem. Res. 32 (1993) 71. [88] Y.-N. Wang, W.-P. Ma, Y.-J. Lu, J. Yang, Y.-Y. Xu, H.-W. Xiang, Y.-W. Li, Y.-L. Zhao, B.-J. Zhang, Fuel 82 (2003) 195. [89] I. Iliuta, F. Larachi, J. Anfray, N. Dromard, D. Schweich, Ind. Eng. Chem. Res. 47 (2008) 3861. [90] N. Hua, H. Wang, Y. Du, M. Shen, P. Yang, Catal. Commun. 6 (2005) 491. [91] D.G. Rethwisch, J.A. Dumesic, J. Catal. 101 (1986) 35. [92] J. Xu, G.F. Froment, AIChE J. 35 (1989) 88. [93] K.M.V. Bussche, G.F. Froment, J. Catal. 161 (1996) 1. [94] M.E. Dry, Product R&D 15 (1976) 282. [95] W.H. Zimmerman, D.B. Bukur, Can. J. Chem. Eng. 68 (1990) 292. [96] W.S.J.Z.B. Zhang, J. Nat. Gas Chem. 5 (1996) 107. [97] G.P. Van Der Laan, A. Beenackers, Catal. Rev. 41 (1999) 255. [98] M. Rahmati, M. Mehdi, M. Bargah-Soleimani, Can. J. Chem. Eng. 79 (2001) 800. [99] A.P. Raje, B.H. Davis, Catal. Today 36 (1997) 335. [100] J.R. Inga, B.I. Morsi, Energy Fuels 10 (1996) 566. [101] B. Jager, R. Espinoza, Catal. Today 23 (1995) 17. [102] J.C.W. Kuo, Two-stage process for conversion of synthesis gas to high quality transportation fuels. Final report, p. Medium: X; Size: Pages: 369 (1985). [103] M. Abbasi, M. Farniaei, M.R. Rahimpour, A. Shariati, Energy Fuels 27 (2013) 5359. [104] T. Boeltken, M. Belimov, P. Pfeifer, T. Peters, R. Bredesen, R. Dittmeyer, Chem. Eng. Process.: Process Intensif. 67 (2013) 136. [105] L. Roses, F. Gallucci, G. Manzolini, S. Campanari, M. van Sint Annaland, Chem. Eng. J. 171 (2011) 1415. [106] T. Graham, Philos. Trans. R. Soc. Lond. 156 (1866) 399. [107] B. Lukyanov, D. Andreev, V. Parmon, Chem. Eng. J. 154 (2009) 258. [108] A.K. Mostafazadeh, M.R. Rahimpour, Chem. Eng. Process.: Process Intensif. 48 (2009) 683. [109] L. Cornaglia, J. Múnera, E. Lombardo, Int. J. Hydrogen Energy 40 (2015) 3423. [110] M.R. Rahimpour, H. Elekaei, Fuel Process. Technol. 90 (2009) 747. [111] K.E. Coulter, J.D. Way, S.K. Gade, S. Chaudhari, G.O. Alptekin, S.J. DeVoss, S.N. Paglieri, B. Pledger, J. Membr. Sci. 405 (2012) 11. [112] R. Dittmeyer, V. Höllein, K. Daub, J. Mol. Cat. A: Chem. 173 (2001) 135. [113] Y. Chen, A.S. Mahecha-Botero, C.J. Lim, J.R. Grace, J. Zhang, Y. Zhao, C. Zheng, Ind. Eng. Chem. Res. 53 (2014) 6230. [114] A. Iulianelli, A. Basile, Catal. Sci. Technol. 1 (2011) 366. [115] A. Basile, F. Gallucci, S. Tosti, Membrane Sci. Technol. 13 (2008) 255. [116] M. Oertel, J. Schmitz, W. Weirich, D. Jendryssek-Neumann, R. Schulten, Chem. Eng. Technol. 10 (1987) 248. [117] D. Mendes, A. Mendes, L. Madeira, A. Iulianelli, J. Sousa, A. Basile, Asia-Pacific J. Chem. Eng. 5 (2010) 111. [118] S. Tosti, A. Basile, G. Chiappetta, C. Rizzello, V. Violante, Chem. Eng. J. 93 (2003) 23. [119] S. Tosti, C. Rizzello, F. Borgognoni, N. Ghirelli, A. Santucci, P. Trabuc, Fusion Eng. Des. 86 (2011) 2180. [120] S. Saeidi, N.A.S. Amin, M.R. Rahimpour, J. CO2 Util. 5 (2014) 66. [121] A.A. Forghani, H. Elekaei, M.R. Rahimpour, Int. J. Hydrogen Energy 34 (2009) 3965.

[122] J.W. Burroughs, J.J.M. Gray, R.L. Herbst, W.C. Moyer, Hydrogenation reactors with improved flow distribution, Google Patents, (1975). [123] D. Mendes, S. Sá, S. Tosti, J.M. Sousa, L.M. Madeira, A. Mendes, Chem. Eng. Sci. 66 (2011) 2356. [124] L. Brandão, L.M. Madeira, A.M. Mendes, Chem. Eng. Sci. 62 (2007) 6768. [125] G. Chiappetta, G. Clarizia, E. Drioli, Chem. Eng. J. 136 (2008) 373. [126] E. McLeary, J. Jansen, F. Kapteijn, Microporous Mesoporous Mater. 90 (2006) 198. [127] M. Vospernik, A. Pintar, G. Ber9 ci9 c, J. Levec, J. Membr. Sci. 223 (2003) 157. [128] G. Sznejer, M. Sheintuch, Chem. Eng. Sci. 59 (2004) 2013. [129] S. Ge, C. Liu, L. Wang, Chem. Eng. J. 84 (2001) 497. [130] O. Schramm, A. Seidel-Morgenstern, Chem. Eng. Sci. 54 (1999) 1447. [131] M.R. Rahimpour, A. Mirvakili, J. Ind. Eng. Chem. 19 (2013) 508. [132] R. Chein, Y. Chen, J. Chung, J. Membr. Sci. 475 (2015) 193. [133] M.R. Rahimpour, M. Dehnavi, F. Allahgholipour, D. Iranshahi, S. Jokar, Appl. Energy 99 (2012) 496. [134] M.M. Mardanpour, R. Sadeghi, M.R. Ehsani, M.N. Esfahany, J. Ind. Eng. Chem. 18 (2012) 1157. [135] A. Brunetti, A. Caravella, G. Barbieri, E. Drioli, J. Membr. Sci. 306 (2007) 329. [136] M. Halabi, M. De Croon, J. Van Der Schaaf, P. Cobden, J. Schouten, Appl. Catal. A: Gen. 389 (2010) 80. [137] A. Adris, C. Lim, J. Grace, Chem. Eng. Sci. 52 (1997) 1609. [138] A. Adris, C.J. Lim, J. Grace, Chem. Eng. Sci. 49 (1994) 5833. [139] J.-P.D.-I.S. Heinrich, Otto-von-Guericke-Universität Magdeburg, Universitätsbibliothek, (2006). [140] S.A.R.K. Deshmukh, J.A. Laverman, A.H.G. Cents, M. van Sint Annaland, J.A.M. Kuipers, Ind. Eng. Chem. Res. 44 (2005) 5955. [141] S. Tosti, Int. J. Hydrogen Energy 35 (2010) 12650. [142] K. Hashi, K. Ishikawa, T. Matsuda, K. Aoki, J. Alloys Compd. 368 (2004) 215. [143] G. Lu, J.D. Da Costa, M. Duke, S. Giessler, R. Socolow, R. Williams, T. Kreutz, J. Colloid Interf. Sci. 314 (2007) 589. [144] S.A.S. Rezai, Y. Traa, J. Membr. Sci. 319 (2008) 279. [145] A. Basile, P. Pinacci, S. Tosti, M. De Falco, C. Evangelisti, T. Longo, S. Liguori, A. Iulianelli, Water Gas Shift Reaction in Pd-based Membrane Reactors, Trans Tech Publ, 2010, pp. 99. [146] A. Basile, F. Gallucci, A. Iulianelli, F. Borgognoni, S. Tosti, J. Membr. Sci. 311 (2008) 46. [147] J.R. Lattner, M.P. Harold, Appl. Catal. B: Environ. 56 (2005) 149. [148] A. Shigarov, V. Meshcheryakov, V. Kirillov, Theor. Found. Chem. Eng. 45 (2011) 595. [149] A. Iulianelli, A. Basile, Int. J. Hydrogen Energy 35 (2010) 3170. [150] C.-Y. Yu, D.-W. Lee, S.-J. Park, K.-Y. Lee, K.-H. Lee, Int. J. Hydrogen Energy 34 (2009) 2947. [151] A. Basile, S. Curcio, G. Bagnato, S. Liguori, S. Jokar, A. Iulianelli, Int. J. Hydrogen Energy 40 (2015) 5897. [152] A. Iulianelli, S. Liguori, J. Wilcox, A. Basile, Catal. Rev. 58 (2016) 1. [153] Y. Lin, Sep. Purif. Technol. 25 (2001) 39. [154] S. Uemiya, Top. Catal. 29 (2004) 79. [155] M.R. Rahimpour, M. Bayat, Ind. Eng. Chem. Res. 49 (2009) 472. [156] M. Sjardin, K.J. Damen, A.P.C. Faaij, Energy. 31 (2006) 2523. [158] T. Sato, H. Yokoyama, H. Miki, N. Itoh, Kagaku Kogaku Ronbunshu 32 (2006) 274. [159] K. Kusakabe, S. Fumio, T. Eda, M. Oda, K.-I. Sotowa, Int. J. Hydrogen Energy 30 (2005) 989. [160] N. Itoh, E. Tamura, S. Hara, T. Takahashi, A. Shono, K. Satoh, T. Namba, Catal. Today 82 (2003) 119. [161] T. Sato, H. Yokoyama, H. Miki, N. Itoh, J. Membr. Sci. 289 (2007) 97. [162] S. Soltani, M. Sahimi, T. Tsotsis, Encycl. Membrane Sci. Technol. (2013). [163] F. Rusting, P.P.A.C. Pex, J.A.J. Peters, G. De Jong, in Sealing socket and method for arranging a sealing socket to a tube, Google Patents (2006). [164] S. Paglieri, J. Way, Sep. Purif. Rev. 31 (2002) 1. [165] L. Mazzone, F. Fernandes, Latin Am. Appl. Res. 36 (2006) 141. [166] M.A. Marvast, M. Sohrabi, S. Zarrinpashne, G. Baghmisheh, Chem. Eng. Technol. 28 (2005) 78. [167] M.J. Keyser, R.C. Everson, R.L. Espinoza, Ind. Eng. Chem. Res. 39 (2000) 48. [168] Y.-N. Wang, Y.-Y. Xu, H.-W. Xiang, Y.-W. Li, B.-J. Zhang, Ind. Eng. Chem. Res. 40 (2001) 4324. [169] S.M.H. Jayhooni, A. Mirvakili, M.R. Rahimpour, J. Nat. Gas Sci. Eng. 9 (2012) 172. [170] M.R. Rahimpour, A. Mirvakili, K. Paymooni, Energy 36 (2011) 1223. [171] M.R. Rahimpour, A.M. Bahmanpour, Appl. Energy 88 (2011) 2027. [172] A. Ghareghashi, S. Ghader, H. Hashemipour, J. Ind. Eng. Chem. 19 (2013) 1811. [173] M.R. Rahimpour, A. Mirvakili, K. Paymooni, Int. J. Hydrogen Energy. 36 (2011) 4917. [174] J. Tong, C. Su, K. Kuraoka, H. Suda, Y. Matsumura, J. Membr. Sci. 269 (2006) 101. [175] T. van Herwijnen, W.A. de Jong, J. Catal. 63 (1980) 83. [176] C. Wheeler, A. Jhalani, E.J. Klein, S. Tummala, L.D. Schmidt, J. Catal. 223 (2004) 191. [177] R.L. Keiski, T. Salmi, V.J. Pohjola, Chem. Eng. J. 48 (1992) 17. [178] S.S. Hla, D. Park, G.J. Duffy, J.H. Edwards, D.G. Roberts, A. Ilyushechkin, L.D. Morpeth, T. Nguyen, Chem. Eng. J. 146 (2009) 148. [179] A.A. Phatak, N. Koryabkina, S. Rai, J.L. Ratts, W. Ruettinger, R.J. Farrauto, G.E. Blau, W.N. Delgass, F.H. Ribeiro, Catal. Today 123 (2007) 224. [180] Y. Li, Q. Fu, M. Flytzani-Stephanopoulos, Appl. Catal. B: Environ. 27 (2000) 179.

S. Saeidi et al. / Journal of Industrial and Engineering Chemistry 49 (2017) 1–25 [181] Y. Choi, H.G. Stenger, J. Power Sources. 124 (2003) 432. [182] N.A. Koryabkina, A.A. Phatak, W.F. Ruettinger, R.J. Farrauto, F.H. Ribeiro, J. Catal. 217 (2003) 233. [183] C.P.P. Singh, D.N. Saraf, Ind. Eng. Chem. Process Design Dev. 16 (1977) 313. [185] W.-H. Chen, M.-R. Lin, T.L. Jiang, M.-H. Chen, Int. J. Hydrogen Energy 33 (2008) 6644. [186] J. Boon, E. Dijk, Ö. Pirgon-Galin, W. Haije, R. Brink, Catal. Lett. 131 (2009) 406. [187] A. Criscuoli, A. Basile, E. Drioli, Catal. Today 56 (2000) 53. [188] Y.-S. Seo, D.-J. Seo, Y.-T. Seo, W.-L. Yoon, J. Power Sources 161 (2006) 1208. [189] O.L. Ding, S.H. Chan, Int. J. Hydrogen Energy 33 (2008) 4325. [190] B. Sarup, B.W. Wojciechowski, Can. J. Chem. Eng. 67 (1989) 62. [191] M.R. Rahimpour, M. Bayat, Int. J. Hydrogen Energy 36 (2011) 6616. [192] M. Khademi, A. Jahanmiri, M.R. Rahimpour, Int. J. Hydrogen Energy 34 (2009) 5091. [193] P. Parvasi, A.K. Mostafazadeh, M.R. Rahimpour, Int. J. Hydrogen Energy 34 (2009) 3717. [194] M. Miyamoto, C. Hayakawa, K. Kamata, M. Arakawa, S. Uemiya, Int. J. Hydrogen Energy 36 (2011) 7771. [195] M.R. Rahimpour, A. Asgari, J. Hazard. Mater. 153 (2008) 557. [196] A. Brunetti, G. Barbieri, E. Drioli, Chem. Eng. Sci. 64 (2009) 3448. [197] K. Sato, S. Hamakawa, M. Natsui, M. Nishioka, T. Inoue, F. Mizukami, Catal. Today. 156 (2010) 276. [198] A. Santucci, M.C. Annesini, F. Borgognoni, L. Marrelli, M. Rega, S. Tosti, Int. J. Hydrogen Energy 36 (2011) 1503. [199] F.A. Silva, C.E. Hori, A.M. da Silva, L.V. Mattos, J. Múnera, L. Cornaglia, F.B. Noronha, E. Lombardo, Catal. Today 193 (2012) 64. [200] A. Basile, P. Pinacci, A. Iulianelli, M. Broglia, F. Drago, S. Liguori, T. Longo, V. Calabrò, Int. J. Hydrogen Energy 36 (2011) 2029. [201] D. Mendes, S. Tosti, F. Borgognoni, A. Mendes, L.M. Madeira, Catal. Today 156 (2010) 107. [202] W.-H. Lin, Y.-C. Liu, H.-F. Chang, J. Chin. Inst. Chem. Eng. 39 (2008) 435. [203] M. Sheintuch, R.M. Dessau, Chem. Eng. Sci. 51 (1996) 535. [204] M.R. Rahimpour, S. Ghader, Chem. Eng. Process.: Process Intensif. 43 (2004) 1181. [205] M.R. Rahimpour, H. Mottaghi, M. Barmaki, Fuel Process. Technol. 91 (2010) 600. [206] G. Barbieri, A. Brunetti, G. Tricoli, E. Drioli, J. Power Sources 182 (2008) 160. [207] M. Abashar, A. Al-Rabiah, Chem. Eng. Processi.: Process Intensif. 44 (2005) 1188. [208] P. Ferreira-Aparicio, M. Benito, K. Kouachi, S. Menad, J. Catal. 231 (2005) 331. [209] F. Gallucci, M. De Falco, S. Tosti, L. Marrelli, A. Basile, Int. J. Hydrogen Energy 32 (2007) 4052. [210] S. Tosti, F. Borgognoni, A. Santucci, Int. J. Hydrogen Energy 35 (2010) 11470. [211] P. Bernardo, G. Barbieri, E. Drioli, Chem. Eng. Sci. 65 (2010) 1159.

[212] [213] [214] [215] [216] [217] [218] [219] [220] [221] [222] [223] [224] [225] [226] [227] [228] [229] [230] [231] [232] [233] [234] [235] [236] [237] [238] [239] [240] [241] [242] [243] [244]

25

M.R. Rahimpour, S. Ghader, Chem. Eng. Technol. 26 (2003) 902. J. Zaman, A. Chakma, Int. J. Hydrogen Energy 20 (1995) 21. E. Gobina, K. Hou, R. Hughes, Chem. Eng. Sci. 50 (1995) 2311. A. Basile, G. Chiappetta, S. Tosti, V. Violante, Sep. Purif. Technol. 25 (2001) 549. A. Basile, L. Paturzo, A. Vazzana, Chem. Eng. J. 93 (2003) 31. K. Sato, T.-a. Hanaoka, S.-i. Niwa, C. Stefan, T. Namba, F. Mizukami, Catal. Today 104 (2005) 260. A. Basile, S. Tosti, G. Capannelli, G. Vitulli, A. Iulianelli, F. Gallucci, E. Drioli, Catal. Today 118 (2006) 237. O. Iyoha, R. Enick, R. Killmeyer, B. Howard, B. Morreale, M. Ciocco, J. Membr. Sci. 298 (2007) 14. N. Itoh, S. Watanabe, K. Kawasoe, T. Sato, T. Tsuji, Desalination 234 (2008) 261. G. Manzolini, S. Tosti, Int. J. Hydrogen Energy 33 (2008) 5571. S.-H. Cheng, H.-J. Chen, H. Chang, C.-K. Chang, Y.-M. Chen, Chem. Eng. Sci. 63 (2008) 1428. A. Caravella, F.P. Di Maio, A. Di Renzo, J. Membr. Sci. 321 (2008) 209. A. Li, C.J. Lim, J.R. Grace, Chem. Eng. J. 138 (2008) 452. L. Bortolotto, R. Dittmeyer, Sep. Purif. Technol. 73 (2010) 51. A. Iulianelli, S. Liguori, V. Calabrò, P. Pinacci, A. Basile, Int. J. Hydrogen Energy 35 (2010) 12626. F. Gallucci, M.V.S. Annaland, J. Kuipers, Int. J. Hydrogen Energy 35 (2010) 1659. A. Di Carlo, A. Dell’Era, Z. Del Prete, Int. J. Hydrogen Energy 36 (2011) 11815. D. Iranshahi, E. Pourazadi, K. Paymooni, M.R. Rahimpour, A. Jahanmiri, B. Moghtaderi, Chem. Eng. J. 178 (2011) 264. R. Koc, N.K. Kazantzis, Y.H. Ma, Int. J. Hydrogen Energy 36 (2011) 4934. C. Yu, H. Xu, Sep. Purif. Technol. 78 (2011) 249. M.R. Rahimpour, S. Mazinani, B. Vaferi, M. Baktash, Appl. Energ. 88 (2011) 41. R. Koc, N.K. Kazantzis, Y.H. Ma, J. Loss Prev. Process Ind. 24 (2011) 852. F. García-García, L. Torrente-Murciano, D. Chadwick, K. Li, J. Membr. Sci. 405 (2012) 30. W. Liang, R. Hughes, Catal. Today 104 (2005) 238. A. Li, W. Liang, R. Hughes, J. Membr. Sci. 149 (1998) 259. P.P. Mardilovich, Y. She, Y.H. Ma, M.H. Rei, AIChE J. 44 (1998) 310. S. Uemiya, Y. Kude, K. Sugino, N. Sato, T. Matsuda, E. Kikuchi, Chem. Lett. 17 (1988) 1687. G. Xomeritakis, Y. Lin, J. Membr. Sci. 120 (1996) 261. M.C. Iliuta, B.P. Grandjean, F. Larachi, Ind. Eng. Chem. Res. 42 (2003) 323. F. Roa, J.D. Way, Ind. Eng. Chem. Res. 42 (2003) 5827. F. Gallucci, L. Paturzo, A. Famà, A. Basile, Ind. Eng. Chem. Res. 43 (2004) 928. J. Tong, L. Su, Y. Kashima, R. Shirai, H. Suda, Y. Matsumura, Ind. Eng. Chem. Res. 45 (2006) 648. J. Tong, Y. Matsumura, H. Suda, K. Haraya, Sep. Purif. Technol. 46 (2005) 1.